UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM ERASMUS UNIVERSITY ROTTERDAM INSTITUTE OF SOCIAL STUDIES THE NETHERLANDS VIETNAM – NETHERLANDS PROGRAMME FOR M A IN DEVELOPMANT ECONOMICS THE EFFECT OF[.]
INTRODUCTION
Research Background
Research has extensively examined the fragmentation of property, land accumulation, and productivity, highlighting the complex relationships among these variables Various studies have revealed inconsistencies regarding land characteristics such as fragmentation, soil fertility, and yield, indicating differing performances across different contexts.
Previous studies in Vietnam have explored land fragmentation and consolidation (Pham et al., 2007; Tran and Vu, 2019; Markussen et al., 2016; Nguyen et al., 2020) The World Bank (2016) highlighted a direct link between land fragmentation and reduced agricultural production Researchers found that by shifting to more profitable crops and increasing machinery use, farmers could lower labor demands and allocate more time to non-farming activities They identified a correlation between labor input in agriculture and land fragmentation However, no significant connection was found between land fragmentation and employment in non-agricultural sectors, indicating potential underdevelopment in rural labor markets.
Agricultural produce, including cultivation, livestock, and aquaculture, serves as the primary income source for rural residents, suggesting that land fragmentation may negatively impact agricultural income and overall household earnings This study computed an index for land fragmentation and measured agricultural income to explore the relationship between these factors The findings confirmed a statistically significant correlation, indicating that land fragmentation adversely affects production outcomes Additionally, the discussion highlighted other relevant aspects related to this issue.
Research Objectives
In rural Vietnam, farming and animal husbandry are the primary sources of household income While previous studies have explored the relationship between land fragmentation and agricultural productivity, they often overlook the direct impact on per capita income, a key indicator of quality of life This study aims to evaluate the connection between farm household income and land fragmentation across various regions in Vietnam, while also considering other socioeconomic factors By utilizing a recent data set, the research highlights the direct effects of land fragmentation on income, which is crucial for understanding agricultural production outcomes Additionally, the study investigates how demographic and educational characteristics significantly influence household income.
Scope of study
Numerous studies have explored the impact of land fragmentation on poverty, income, and production levels This research aims to utilize a specialized dataset on rural Vietnam to generate actionable insights for rural development, particularly in relation to land use policies that influence demographics and educational attainment The findings will contribute directly to enhancing rural development efforts in Vietnam.
Research Structure
This article outlines the structure of the research paper, beginning with Chapter 1 – Introduction, which presents the research background and scope of study Chapter 2 – Literature Review discusses theories and findings related to land fragmentation and its impacts Chapter 3 details the study methods and data utilized, while Chapter 4 presents the results of the analyses Finally, Chapter 5 concludes with the findings, limitations of the study, and recommendations for future research.
LITERATURE REVIEW
Definitions
Land is a limited resource essential for livelihood and financial security, often inherited and converted into wealth (Ellis, 1992) Hartvigsen (2014) identifies two key aspects of agricultural land allocation: ownership dispersion and land use dispersion In both developed and developing countries, agricultural production and the livelihoods of many depend heavily on land This connection drives interest among scholars and policymakers in the efficient distribution and utilization of land resources, highlighting the importance of measuring and analyzing land fragmentation.
The term, "fragmentation", refers to a splitting up of a previously integrated production process into two or more components, or "fragments" (Jones R W.,2000)
Land fragmentation refers to a scenario where a farm is made up of multiple parcels of land that are not contiguous (Binns, 1950; King and Burton, 1982; Blarel et al., 1992; Pham et al., 2007) According to McPherson (1982), it involves a farmer cultivating several spatially separated plots of either owned or rented land (cited in Veljanoska, 2018).
Land fragmentation, as defined by Van Dijk (2003), involves both the physical division of land parcels and the legal claims associated with them, representing two distinct layers This phenomenon can be assessed on various scales, which determine what is considered "the whole." The research identifies four types of land fragmentation: (1) fragmentation of land ownership, (2) fragmentation of land use, (3) fragmentation within a farm, and (4) the separation of ownership and use While often viewed negatively, land fragmentation can offer several advantages, including ecological and aesthetic benefits, as well as a reduced risk of crop loss due to disease.
4 extreme weather Some types of farming also need spatially separated parcels so they can be used for various agricultural purposes
Land fragmentation occurs when a farmer possesses a large amount of disconnected land spread over a wide area, which is common in many countries and often hinders productivity and modernization (Sundqvist and Andersson, 2007) This phenomenon is particularly prevalent in developing nations, where fragmented land can have multiple meanings and implications for agriculture (Pham et al., 2007) The varying quality of these plots allows farmers to diversify their crops, manage labor demands, and mitigate risks related to output and pricing (Ciaian et al., 2018; Tran and).
Land fragmentation refers to a single farm comprising multiple spatially scattered parcels of land This occurs when a family possesses several non-contiguous pieces of land spread across a large area The phenomenon of land fragmentation presents both costs and benefits for agricultural productivity, with its effects varying based on specific circumstances.
Land fragmentation indices
Land fragmentation is a geographical phenomenon shaped by several factors, including organizational size, the number of land parcels owned, the size and shape of each parcel, as well as their spatial and size distribution, as highlighted by King and Burton (1982).
The average number of land parcels held by farmers in a region or country serves as a key indicator of land fragmentation This evaluation is based on two main characteristics: the average size of the holdings and the average parcel size.
In 1961, a method for measuring land fragmentation was proposed, focusing on the percentage of a property owner's land that is not contiguous with the farm Building on this, Simmons (1964) introduced a land fragmentation index that considers both the number of land parcels owned and their respective sizes The calculation of this index is based on these two key factors.
The fragmentation index (FI) is defined by the formula where "n" represents the number of land parcels owned, "a" denotes the size of each land parcel, and "A" signifies the total number of holdings An FI value of 1 indicates a specific level of land fragmentation.
5 that the holding consists of a single parcel of land, while values close to 0 indicate greater fragmentation
Dovring (1965) evaluated land fragmentation by calculating the distance farmers must travel to reach each of their plots, considering the need to return to their farm after each visit (cited in Kadigi et al.).
The K-index, proposed by Januszewski (1968), quantifies land fragmentation by calculating the ratio of the number of land parcels to the distribution of parcel sizes This metric provides a more comprehensive understanding of land use, as it considers the frequency of trips made annually and the possibility of accessing various property areas without the need to return to the farm.
The K-index, which ranges from 0 to 1, indicates the degree of land fragmentation, with values approaching zero signifying significant fragmentation It has three key characteristics: first, fragmentation rises with an increasing number of land parcels; second, fragmentation intensifies as the size range of smaller parcels expands; and third, fragmentation diminishes when the area of larger parcels increases while the area of smaller parcels decreases.
Igozurike (1974) proposed a relative index for measuring land fragmentation, which considers both the average size of agricultural plots and the distance a farmer travels to visit each plot in a round trip The index is computed using a specific formula that reflects these factors.
The fragmentation index (P i) of holding i is calculated using the size of each land parcel (S i) and the total round-trip lengths of all parcels (D t) However, the P-index has limitations, as it lacks a clear definition of distance and does not consider the number of land parcels involved (King and Burton, 1982).
The P 0 land fragmentation metric, introduced by Schmook in 1976, quantifies the ratio of the area of a polygon that encompasses all plots owned by farmers to the area of the main farm This index consistently yields values greater than one, with higher P 0 values signifying greater land fragmentation.
Furthermore, the Simpson index is commonly utilized in the land fragmentation literature (Blarel et al., 1992, Pham et al., 2007; Latruffe and Piet, 2014; Ciaian et al., 2018;
The Simpson index, as noted by Tran and Vu (2019), is sensitive to both the size of the parcel and the number of portions of land It is calculated using a specific formula.
The Simpson index (SI) quantifies land concentration, where \( A_i \) represents the area of the \( i \)th parcel and \( A \) denotes the total farm area A value of zero indicates a concentrated landholding with a single parcel, while a value close to one suggests that a family possesses multiple highly fragmented plots of land (Pham et al., 2007).
Effects of land fragmentation
Land fragmentation significantly impacts agricultural productivity, as explained by classical economic theory (Hartvigsen et al., 2014) Ricardian principles highlight the law of diminishing returns, where fixed land resources lead to declining labor productivity, especially on less fertile soils As cultivation expands into poorer terrains due to capital accumulation, production costs increase To optimize earnings, farmers must effectively combine commercial and natural inputs, while Ricardo's framework emphasizes the roles of land and labor, often neglecting technological advancements.
The Economies of Scale perspective highlights that producers gain cost advantages by expanding production, particularly in agriculture, where larger land areas facilitate mechanization, irrigation, and organized commodity production, as opposed to fragmented land Additionally, Production Theory emphasizes that producers select resource combinations—such as capital, labor, technology, and natural conditions—to maximize profits and enhance production efficiency This underscores the importance of agricultural resources, including natural resources, the environment, and biodiversity, with land being a vital element that serves as both a labor object and an investment, as labor in agricultural production cannot be easily substituted.
The Cobb-Douglas production function or Stochastic frontier Approach, according to the production theory-based explanation, is frequently used to estimate the influence of land
7 fragmentation on output and production efficiency (e.g., Pham et al., 2007; DiFalco et al., 2010; Manjunatha et al., 2013; Deininger et al., 2017; Tran and Vu, 2019)
Tornado (1985) outlines a three-stage model of agricultural development, beginning with subsistence agriculture, where production relies on soil fertility and minimal capital investment, resulting in internal consumption and diminishing profits as production extends to less fertile land The second stage involves agricultural restructuring towards diversification, incorporating new plant and animal varieties, chemical fertilizers, and irrigation to enhance productivity and cater to market demands In the final stage, agriculture advances towards modernity, characterized by specialized farms where capital and technology play crucial roles in production efficiency, leading to economies of scale and a complete supply of technical outputs to the consumer market.
Soil fragmentation plays a crucial role in alleviating the effects of crop failures caused by drought, hail, epidemics, and other natural disasters, thereby helping to reduce hunger This is particularly relevant in diverse farming environments, where varying soil types and growing conditions exist While fragmented soils may be less effective in developed regions like China, Vietnam, Bangladesh, and Europe, these areas continue to support modern, market-oriented agricultural practices.
A study by Mwebaza R (2002) revealed that Ugandan farmers perceive land fragmentation as having both advantages and disadvantages The farmers highlighted the benefit of cultivating diverse crops due to the varying soil fertility However, they also pointed out the challenges of managing scattered land holdings and the time lost in traveling between different plots.
Economic theories and empirical evidence provide a strong basis for studying the impact of land fragmentation on agricultural production in specific local contexts Among the various production organizations globally, the household economy and farm economy are the most prevalent.
2.3.1 The effect of land fragmentation on agricultural production
Many developing nations are grappling with land fragmentation due to household settlement patterns, population growth, and cultural factors, primarily stemming from generational inheritance divisions (Simons, 1985; Niroula and Thapa, 2005; Pham et al., 2007) Simons (1985) highlighted that farming on fragmented land is inefficient, leading to increased production costs and challenges in disease control due to reliance on neighboring farms Rembold (2003) further illustrated the detrimental effects of land fragmentation on agriculture in Central and Eastern Europe and the Commonwealth of Independent States, where average farm sizes range from 0.5 to 2.5 hectares, hindering the adoption of new production models and technology, ultimately resulting in decreased agricultural productivity and efficiency.
Fragmented land, as highlighted by Mwebaza and Gaynor (2002), poses significant challenges for large-scale agricultural activities in Uganda, hindering the development of essential infrastructure such as transportation and irrigation systems Similarly, Niroula and Thapa (2005) emphasize that land fragmentation not only accelerates deterioration and stifles agricultural progress but also discourages collaboration among farmers in South Asia, ultimately leading to increased costs and reduced output.
Land fragmentation in Bangladesh adversely impacts the productivity and technical efficiency of the rice production industry, as noted by Rahman and Rahman (2009) Farmers who own their land tend to achieve greater financial success compared to those who rent land, primarily due to the diverse soil types that increase production costs Furthermore, this fragmentation hinders the country's ability to adopt modern technologies and contributes to a lack of uniformity in agricultural practices.
According from He, M (2014) to analysis of the impact of land fragmentation on agricultural production cost, China has a huge number of people but not much land that can
Land fragmentation can have both positive and negative effects on agricultural production Research in Northwest China indicates that while land fragmentation negatively impacts the total production cost per unit of wheat in some irrigation districts, it does not significantly affect costs in others Smaller farm sizes often lead to lower agricultural productivity due to increased labor reliance and limited mechanization, as noted by Markussen et al (2016) Ciaian et al (2018) further highlight that smaller-scale production tends to employ more labor, which can hinder efficiency In contrast, fragmented land in Albania promotes diversification among household farmers, while Deininger et al (2017) show that in India, it raises wheat cultivation costs, favoring labor over machinery This situation can benefit self-sufficient households but poses challenges for market-oriented farmers who require larger-scale production However, Blarel et al (1992) argue that land fragmentation can also enhance risk management for farmers, offering benefits such as increased farming diversity and reduced labor shortages during harvests, as evidenced by studies in Ghana and Rwanda.
The 1992 study challenged the assumption that fragmentation leads to inefficiency and decreased yield, revealing that consolidation may not significantly enhance land productivity It also found that the private benefits of fragmentation could equal or surpass its costs Consequently, the author recommended that policymakers deepen their understanding of these dynamics.
Fragmentation in markets for property, labor, credit, and food leads to inefficiencies Research following Tan et al (2010) in China indicates that land fragmentation significantly affects economic costs However, Ali et al (2019) found that fragmented land can provide benefits by reducing production shocks and enhancing productivity without raising costs or diminishing agricultural efficiency Despite these potential advantages, efforts to tackle land fragmentation in Rwanda may face challenges.
Fragmented land serves as an effective alternative for mitigating weather hazards, particularly in regions with stagnant loan and insurance markets, as demonstrated by Veljanoska (2018) In Uganda, studies indicate that such land fragmentation can significantly reduce crop yield losses due to erratic rainfall patterns Additionally, Tan et al (2010) highlight that fragmented land is crucial for enhancing the technical efficiency of early rice production, with productivity in South-East China positively correlating with the number of farmers' land This adaptability of fragmented land within industrial organization is further supported by Blarel et al (1992) Moreover, Kadigi et al (2017) found that in Tanzania, fragmented land positively impacts agricultural productivity, suggesting that the local farming environment is a key determinant of how land fragmentation affects farmers' well-being.
DiFalco et al (2010) highlight that land fragmentation adversely affects the earnings of Bulgarian farmers while encouraging crop diversity This aligns with earlier studies (Simons, 1985; Rembold, 2003; Pham et al., 2007), which indicate that the spatial distribution of fragmented fields complicates cultivation, machinery use, and irrigation Consequently, a decline in farm income emerges as a direct and immediate consequence of land fragmentation on agricultural production.
Research by Latruffe and Piet (2014) highlights that for wheat farmers in France, the impact of land fragmentation on production outcomes is contingent upon the specific usage index employed to measure fragmentation In general, fragmented land tends to increase production costs.
Summary and conceptual framework
This chapter explores various theories and methods related to land fragmentation, highlighting its prevalence in certain developing countries Key factors contributing to land fragmentation include inheritance practices, family settlement models, population growth, land market dynamics, and cultural perspectives Research indicates that fragmented land can negatively impact agricultural production by increasing costs and limiting the use of mechanized equipment, ultimately reducing productivity and efficiency However, some evidence suggests that smaller land parcels may also enhance production under specific conditions The agricultural environment of a region plays a crucial role in determining the effects of land fragmentation on farmers' interests, necessitating scenario-based validation to accurately assess its impact on agricultural productivity.
Household income was chosen as the expected dependent variable because it allowed for more direct measurement of the impact of land fragmentation Because Vietnam is
Vietnam is primarily an agricultural nation, with most rural households relying heavily on agricultural income Evaluating the impact of land fragmentation on household income from agriculture is therefore crucial By treating income as the dependent variable, a production function like the Cobb-Douglas can be employed to assess the effects of land fragmentation Additionally, incorporating various independent variables is essential to comprehensively understand household income from agricultural activities in rural Vietnam.
This study aims to investigate the impact of fragmented landholdings on rural income, while also considering other influential factors such as ethnicity, age, marital status, livestock activities, working-age population, household size, and land types owned by households This objective arises from conflicting findings in previous research, where some studies suggest that land fragmentation negatively affects household income, whereas others propose that factors like work skills, education level, and social networks have become more significant than land ownership in determining rural income.
RESEARCH METHODOLOGY
The situation of land fragmentation in Vietnam
Vietnam has a total of 12,388 million hectares of land suitable for agriculture, primarily managed by state-owned enterprises for perennial crop production Approximately 9.6 million farming households utilize the remaining land, averaging 0.8 hectares each, often fragmented into four non-contiguous plots While large agricultural tracts constitute only about 10% of the total land area, they dominate the agricultural landscape Although private corporations manage most farmland, the state retains ownership and control The primary aim of state-managed agricultural land is to produce resilient crops year after year, highlighting the issue of land fragmentation in Vietnam The OECD notes that land acquisition for expanding cattle production has commenced, yet the agricultural sector remains underdeveloped, with few plots exceeding 2 hectares This fragmentation poses a significant challenge to agricultural growth, which is essential to meet the demands of an increasingly discerning consumer market Evidence suggests that agricultural output has declined due to this fragmentation, exacerbated by rapid urbanization over recent decades.
The "2016 Agriculture, Rural and Fisheries Census" by the General Statistics Office of Vietnam reveals that 36% of households utilizing agricultural land possess less than 0.2 hectares Furthermore, households engaged in agricultural activities represent 36% of all households in Vietnam.
5 hectares or more accounts for just over 2%, which drives up production costs, requires a
The fragmentation of land ownership among individual farmer households hinders long-term investments in agriculture, as it leads to dispersed fields and varied planting methods This situation complicates access to essential resources such as transportation, irrigation, and mechanization, while also limiting the adoption of new technologies Consequently, the potential for large-scale, consistent production across different types and strains of agricultural products is significantly restricted.
The government introduced Directive No 10/1998/CT-TTg on February 20, 1998, and Directive No 18/1999/CT-TTg on July 1, 1999, to promote the consolidation of agricultural land for large-scale production Local authorities developed plans to transform small plots into larger, more productive agricultural areas to support farmers (Luu, 2017) However, land accumulation has progressed slowly, with the average number of land parcels per household decreasing from 4.9 in 2006 to 3.9 by 2017 (Luu, 2017; Nguyen et al., 2020) A key policy actively implemented is the concentration of land through new-style cooperatives and large field production models, reflecting significant policy support for land consolidation.
The Land Law (2013) in Vietnam aims to support investment activities by extending agricultural land use terms, reflecting a significant shift in land policy This law promotes effective land utilization through measures that facilitate land accumulation and concentration for modern agricultural production Households can now receive land allotments for up to 50 years, with the possibility of extension for continued production needs However, expanding production often requires the relocation and integration of multiple units to form larger companies and corporations, posing challenges for businesses due to land use limitations.
The upcoming land policy must be revised to establish new and suitable regulations that facilitate legally binding transactions and promote the growth of agricultural businesses.
Since 2013, Vietnam has implemented significant changes to its agricultural land policy, including regulations on business eligibility for agricultural land use rights transfers in 2017, the creation of a pilot organization to promote land concentration through rental price support, and the introduction of tailored regulations for rice fields in 2016 These reforms aim to establish a legal framework that enhances transactions in the agricultural land market and fosters business growth in agricultural production.
Nguyen, H (2014) provides a concise analysis of the benefits and drawbacks of land fragmentation, revealing that the costs associated with land fragmentation surpass its benefits This finding aligns with previous studies and recent government policies advocating for land consolidation.
Table 1 Cost and Benefit of land fragmentation
Costs of land fragmentation Benefits of land fragmentation
Private cost Public cost Private benefit Public cost
• Land loss due to boundaries
• Difficulties in technological application and mechanization
• Delay of mechanization and technological application
• Difficulties in crops planning and land use planning
• Equality of land redistribution (egalitarian principle)
Source: He, M (2014) An analysis of the impact of land fragmentation on agricultural production cost: evidence from farmers in Gansu province, PR China.
Econometric models
Measuring land fragmentation is a complex process that goes beyond merely counting non-contiguous plots or co-owners; it also involves factors such as plot size, shape, distance between plots, and travel time from home to the plot (Latruffe and Piet, 2014) To quantify land fragmentation, many researchers utilize the Simpson's diversification index, which accounts for the number and size of plots as well as the overall size of the farm In this study, the Simpson's index is employed to evaluate land fragmentation, calculated using the formula: \(1 - \left(\frac{\sum a_i^2}{A^2}\right)\) The index ranges from zero to one, where values closer to one indicate higher land fragmentation, suggesting that a household manages multiple plots, while a value of zero signifies complete land consolidation with only one plot.
3.2.2 Estimating the effect of land fragmentation on household income
In agricultural economics, several production functions are utilized, including the Spillman function, the transcendental production function, the Cobb-Douglas function, the de Janvry Modification function, and the polynomial form.
The Cobb-Douglas production function serves as the foundation for various functions aimed at addressing the issue of constant elasticity of input, with the exception of the Spillman function and polynomial forms In this study, the parameters of the inputs remain constant, which justifies the continued use of the Cobb-Douglas production function.
The theoretical foundation of this study is the Cobb-Douglas Production Function, a classic economic theory that elucidates the relationship between input factors and output (Enaami et al., 2013; Prajneshu, 2008) This production function is widely utilized at an economy-wide level and is primarily employed to calculate the production factors for labor.
In 1920, Paul Douglas introduced a production function to demonstrate the relationship between labor and capital Over the years, this function has evolved and been refined, becoming a widely recognized tool for estimating macroeconomic performance across various industries.
The basic Cobb-Douglas production function is specified as below:
Y: the output or the total production K: the capital input (the factor of production) L: the labor input (the factor of production) A: Total Facor Productivity (TFP) A is a positive constant α: the output elasticity of capital β: the output elasticity of labor
The Cobb-Douglas production function exhibits several key characteristics: (i) output elasticity remains constant within a specific industry; (ii) the marginal product is positive but decreasing, indicating diminishing marginal returns, where increases in capital or labor lead to smaller increments in total production; and (iii) returns to scale reflect proportional changes in output when all factors are adjusted equally, with the overall returns to scale being the sum of the output elasticities of capital and labor.
To analyze the effect of land fragmentation on household income, the thesis employed a Cobb–Douglas production function in a double-log format (Ravallion & Van de Walle, 2008) This approach is appropriate as it incorporates various inputs, including age, marital status, and education level, allowing for a comprehensive examination of agricultural household income per capita.
Data
3.3.1 Vietnam Access to Resources Household Survey (VARHS)
This study utilizes the Vietnam Access to Resources Household Survey (VARHS) data, conducted biennially in rural areas across 12 provinces in Vietnam since 2006.
2006 The purpose of this survey was to provide the researchers with the data necessary to determine the effect that land fragmentation has on the income generated from agriculture
Prior to the inaugural pilot of the VARHS in 2002, there was a significant lack of information on how households acquired essential resources like land, credit, and labor Furthermore, there was insufficient understanding of the effectiveness and efficiency of these markets.
The VARHS questionnaire was created to assess the impact of government policies and changes, focusing on household access to productive resources such as land, physical, financial, human, and social capital It aims to uncover the reasons behind restricted resource access for certain households and how they acquire their resources Additionally, the questionnaire addresses various topics, including rural employment, income-generating activities, rural enterprises, property rights, savings, investment, insurance, and participation in both formal and informal social networks.
Since 2006, the VARHS data has been collected from over 2,150 families across 12 provinces in Vietnam Key contributors to this research include the Central Institute for Economic Management (CIEM) and the Institute of Labour Science and Social Affairs (ILSSA), both affiliated with Vietnam's Ministry of Planning and Investment and Ministry of Labour, Invalids, and Social Affairs, respectively.
The VARHS dataset, similar to the Vietnam Household Living Standards Survey (VHLSS), is widely used in research focusing on inequality, poverty, and expenditure analysis related to farmers' economics and living standards, including areas such as education and health care.
The Vietnam Access to Resources Household Survey (VARHS) dataset, introduced in 2002, addresses the need for detailed information on family resources and interactions in Vietnam Since 2006, it has collected data from households across twelve provinces every two years, focusing on how Vietnamese farmers access and utilize production resources, including physical and financial capital The VARHS dataset is instrumental in research related to inequality, poverty, and expenditure analysis, similar to the Vietnam Household Living Standards Survey (VHLSS) This study employs VARHS to examine the effects of land fragmentation on agricultural household income, alongside specific soil-related factors such as soil fertility and irrigation.
This research focused on annual cropland in Vietnam, highlighting the unique characteristics of the country's agricultural system, including seasonal variations and the predominance of annual land, which constitutes over half of the total plots collected at the household level.
Agricultural income calculations have overlooked the revenues and expenses linked to livestock production While livestock contributes to overall agricultural income, its production is largely independent of land-related factors.
This study utilizes data from the VARHS 2018 survey to examine the effects of land fragmentation on agricultural household income, alongside various soil-related factors such as soil fertility and irrigation, as well as land slope The research encompasses a comprehensive analysis of the following aspects: general characteristics of households and dwellings, agricultural land details including disaster information, agricultural practices such as crop cultivation and land transactions, livestock and forestry resources, employment and income sources, extension services, household expenses and savings, credit and social networks, migration status, and the influence of political connections and information sources within rural communities.
The neighborhood's infrastructure encompasses essential elements such as roadways, water sources, power supply, markets, and educational systems Additionally, it is important to consider demographic information about the community, including agricultural aspects like cultivated crops, land sales, rental agreements, land types, and total land area Furthermore, understanding the income and employment status of families, including their primary sources of income and enterprise activities, is crucial for a comprehensive overview of the area.
This study categorizes 12 survey provinces into four distinct regions: Region 1 includes the Northern lowlands, comprising Ha Tay, Phu Tho, and Nghe An; Region 2, the Northern Highlands, consists of Lao Cai, Dien Bien, and Lai Chau; Region 3 covers the Central Highlands with Dak Lak, Dak Nong, and Lam Dong; and Region 4, the Southern Lowlands, includes the delta provinces of Quang Nam, Khanh Hoa, and Long An The analysis reveals significant differences along the north-south and highland-lowland axes, with each region exhibiting substantial disparities, leading to independent analyses for each area.
The VARHS 2018 data on 15,442 acres of land was categorized by ownership type and land function This classification includes various purposes such as annual and perennial crop cultivation, forests, ponds, lakes, grassland, residential properties, and land with buildings and gardens Notably, an estimated 9,201 plots, accounting for 59.58 percent of the total, are designated specifically for the production of annual crops.
Figure 1: Plots per using purpose
This research focuses on specific land ownership modes, which are not considered at this time The data utilized in this article is from the 2018 dataset, encompassing households with annual cropland, including owned, rented, and undocumented land Additionally, agricultural income is calculated without factoring in revenues and expenditures from livestock production, as livestock contributes minimally to land-related issues.
To enhance the reliability of results and draw insightful conclusions, various statistical methods were employed Each response was meticulously coded and verified for accuracy prior to analysis Stata software facilitated both descriptive statistical analysis and econometric model estimation throughout the data processing.
Most of the the research's analyses are carried out on their own accord for each region, and the results show that there are significant differences between them.
Methodology
Most research use Simpson's diversification index to quantify farmland fragmentation, which takes into consideration the number of plots, plot size, and farm size; however, this
Annual crops Perennial crops Land for forests Land for ponds and lakes Grass land, , and other types of land Land with buildings and gardens on land Residential Land
23 study uses The Simpson's index of land fragmentation to assess the index of land fragmentation
The thesis used a Cobb–Douglas production function in the form of a double-log function to model the effects of land fragmentation on household income (Ravallion & Van de Walle, 2008)
As shown in equation, the agricultural household per capita income was expected to be a function of land and other explanatory variables:
LnY represents the natural logarithm of the per capita income derived from agricultural activities within a household The variable X encompasses a range of factors related to the household and its head, including household size, the number of laborers, livestock ownership, as well as the age, education, marital status, and ethnicity of the household head.
Z: is a vector of commune variables of land fragmentation index with intrusment variables e: is the error term
The land fragmentation index may exhibit endogeneity due to its dependence on regional characteristics and household income, leading to skewed and inconsistent estimates when using OLS methodology To address this issue, the instrumental variables (IV) method is recommended for generating consistent estimators (Wooldridge, 2013) It is essential to conduct multiple IV tests to ensure that the instrument relevance and exogeneity assumptions are met, thereby avoiding the use of invalid and weak instruments that yield imprecise estimates For the IV method to be effective, an instrumental variable should be created based on heteroskedasticity rather than the traditional exclusion condition This approach has been favored in numerous studies, successfully resolving the endogeneity problem (Sabia, 2007; Giambona & Schwienbacher, 2008; Huang et al., 2009; Emran and Hou, 2013; Zhao, 2015).
In this study, the IV method was utilized to identify external instruments that were either unavailable or did not meet the necessary criteria (Lewbel, 2012) The analysis revealed that the instrumental variables employed to compute the land fragmentation index were associated with the average slope.
The selection of regions 1 and 2 as instrumental variables is justified by their representation of 71.75 percent of the surveyed households Additionally, the average slope of land plots, categorized into four levels from flat to extremely sloped, serves as another instrumental variable based on geographical characteristics from the VARHS dataset The analysis includes various independent variables related to the household head, such as marital status, ethnicity, education level, and age, as well as household characteristics like size, number of laborers, and livestock production activities Education levels of household heads are classified into three groups: those who completed general education, those who participated in short-term or professional diploma programs, and those with a college diploma or higher Furthermore, the livestock status is represented as a dummy variable indicating whether a household engages in agricultural breeding activities.
To determine the reliability of the instruments used in this study, it is crucial to avoid employing invalid and unreliable tools that yield inaccurate estimates and misleading conclusions (Baum, Schaffer, & Stillman, 2003) The validity of these instruments was assessed using an over-identifying constraints test, and the Hansen J-statistics indicated no statistical significance, suggesting that the instrumental variables are indeed reliable (Baum, Schaffer, & Stillman, 2003).
RESULTS
Descriptive Statisitcs
Variable Obs Mean Std Dev Min Max
Logarit of income per capita
Household Head's Age 2,531 52.7108 12.9518 10 101 Household Head's Marrital
The investigation's key findings are summarized in Table 2, which presents a comprehensive analysis of the dataset comprising 2,531 observations, representing 2,531 unique households The survey results indicate that 47.25 percent of the households belong to the Kinh ethnic group, while 52.75 percent are from other ethnic groups Notably, 17.54 percent of respondents reported no marital status, with the majority of household heads being married, constituting 82.46 percent of the total The average age of homeowners is approximately 52.7 years, and a significant 88.36 percent of respondents have not pursued education beyond primary school, with only 3.82 percent holding a degree from an accredited institution The number of residents per household varies from one to twelve.
26 household that was surveyed had an average of between four and five inhabitants A household could have anywhere from one person to twelve people living there at any given time
Table 3 Mean of variables follow 4 regions
Mean of Land Fragmentation index
Mean of Income from Agriculture
Table 3 highlights the mean values for land fragmentation, agricultural income, household numbers, and household labor across various geographic regions in Vietnam The northern provinces are notably facing high land fragmentation, while the highland provinces exhibit lower household fragmentation rates Additionally, the Central Highlands generate significantly higher agricultural income compared to other regions Households in the Northern Highlands are larger, averaging over five members, whereas other provinces maintain a more consistent household size Interestingly, while the Northern Lowlands, Central Highlands, and Southern Lowlands have similar household labor numbers, the Northern Lowlands show a markedly lower average number of workers per household.
The survey included a diverse range of homes across all four geographic subgroups, with a particular emphasis on the northern provinces of Vietnam Notably, the mountainous northern provinces of Lao Cai, Dien Bien, and Lai Chau contributed 43% of the total responses Interestingly, the number of households participating in the survey from the Central and Southern Regions of Vietnam matched that of the northern mountainous provinces.
(1) Northern lowlands: Ha Tay, Phu Tho, and Nghe An
(2) Northern highlands: Lao Cai, Dien Bien and Lai Chau
(3) Central highlands: Dak Lak, Dak Nong, and Lam Dong
(4) Southern lowlands: Quang Nam, Khanh Hoa, and Long An
N UM BE R O F H O US EH O LD
To quickly determine the number of employed individuals in a household, divide the total family members by 18, providing an accurate count of those employed and under 60 years old A survey of 2,531 homes revealed that most families typically have between two and four working-age members.
Figure 4: Literacy of Household Head
The educational level of the head of the household is determined by the highest degree achieved Data collection is based on this criterion, revealing that 88.34 percent of respondents had not received intensive training Among those who underwent retraining, only 175 heads of households completed short-term training Additionally, 11.64 percent of survey participants reported having received either short-term or medium-term training, or having completed education from intermediate to higher levels.
H ig h se t D ip lo ma
Figure 5: Household size in regions
Figure 5 illustrates the relationship between household workforce size and regional distribution, highlighting the number of homes in each area Notably, in area 2 (the Northern mountainous provinces), there is a significant disparity in households with four to six workers compared to those with fewer or more workers, as this region has a considerably higher number of households within that specific workforce range.
Figure 6: Scatter of Land Fragmentation and Region
N UMB ER O F H O US EH O LD
The scatter plot comparing the land fragmentation index across four regions shows consistent values ranging from 0 to 1 Notably, the southern regions (region 3 and region 4) exhibit significantly less land fragmentation than the northern regions (region 1 and region 2) This disparity may be attributed to the predominant village culture in the northern delta provinces, where high land fragmentation occurs due to the practice of parents dividing family farmland among their children as they reach adulthood.
Figure 7: Scatter of Land fragmentation index and log of agriculture income per capita
The land fragmentation index significantly impacts the dependent variable, as evidenced by a scatter plot illustrating the relationship between the index and the logarithm of average agricultural family income This visual representation confirms the connection between these two variables.
Figure 8: Livestock status and Marital status in regions
The graph illustrates the correlation between the marital status of household heads and livestock production across four regions Livestock production represents a significant portion of household activities in all areas.
The "married" status of households constitutes a significant portion across all regions, indicating a correlation between marriage and livestock production Notably, in region 2, households led by married individuals exhibit the highest percentage of livestock activities This trend can be attributed to marriage fostering complementarity in livestock management, as it increases the workforce and enhances the need for income through livestock endeavors.
N UM BE R O F H O US EH O LD
Livestock status and Marital status
Have livstock Activities Don't Have livestock Activities Household head Married
Figure 9: Scatter of labors in household and log of agriculture income per capita
The scatter plot comparing working-age household members to dependent variables shows a consistent distribution of per capita income across all households, irrespective of employment numbers Notably, households with seven or more employees exhibit a lower average per capita income compared to those with fewer employees.
The scatter plot illustrates the relationship between the age of household heads and the logarithm of agricultural income per capita It reveals that the majority of household heads fall within the age range of 40 to 60 years.
Households led by individuals aged 40 to 60 generate the highest per capita income from agricultural activities, making them the most financially successful group in this sector.
Regression Results
The first step is to perform an autocorrelation test on the variables in the equation:
Marital status of household head
Number of labors in household
Professio nally trained – household head
Marital status of household head
Number of labors in household
Table 4 indicates that all correlation coefficients are below 0.5, with the exception of the correlation between the number of laborers in a household and the size of the household, which stands at 0.7063 Most of these coefficients are significant at the 5% level, suggesting that there is no necessity to eliminate any variables from the models.
Following an analysis of all the data acquired using the approach described in the prior chapter, the table of findings consists of the items listed below:
Marital status of household head 0.3634** 0.0862
Number of observations: 2,530 Prob > chi2 = 0
The findings presented in the table reveal that land fragmentation negatively impacts agricultural income per capita A negative correlation exists between agricultural income and several factors, including the number of employees, the presence of livestock activities, and the ethnicity of the household head Additionally, instrumental variables such as the slope of the land at the household level and the geographical distinctions between region 1 (northern lowlands) and region 2 (northern highlands) are used to describe the distribution of these findings.
A one percentage point increase in the land fragmentation index leads to a 0.031% decrease in the income per capita of agricultural households, aligning with previous research on land fragmentation in Vietnam Additionally, the income disparity between Kinh-headed households and those led by others is projected to be 42.55% under constant conditions, reflecting the characteristics of the data collected from specific provinces.
The analysis reveals a notably higher proportion of ethnic minorities in the region compared to other provinces Projections indicate that farm household income per capita increases by 1.14 percentage points for each additional year of the household head's age, assuming all other factors remain constant Furthermore, households with a married head experience a predicted agricultural household per capita income that is 43.83 percent higher than those with a single head, under the same conditions Engaging in livestock activities significantly boosts per capita income, with agricultural households involved in livestock earning 147% more than those that do not This disparity is likely due to the higher selling prices of livestock-derived products compared to those from farming Additionally, for every extra working-age individual in a household, there is an anticipated 12.6% decrease in per capita income, assuming no changes in other relevant factors Notably, household size and the education level of the household head do not significantly impact per capita income calculations based on the number of employed family members, as these variables are interrelated.
After the regresssion, the result of the Hansen's J Test was used to test the appropriateness of the instrumental variables that were used in this test, was Hansen's J chi2
(2) = 5.71027 with p = 0.0575, were not found to be statistica It means that the instrumental variables (Region 1, Region 2, and the variable measuring the slope of the land) should be considered reliable
The author conducts endogenous tests to assess the exogeneity of the regressors in the model, hypothesizing that these variables are exogenous The results of the analysis are presented accordingly.
Therefore, the land fragmentation, the target independent variable to find the relationship between it and dependent variable is endogenous variables
The first - stage regression summary statistics the results as shown in the table below
Table 6 First-stage regression summary statistics
Land fragmentation negatively impacts household agricultural revenue per worker, as smaller parcels require more labor to cultivate Factors such as the marital status of the household head and their age positively influence the dependent variable in this study Additionally, livestock-related activities have been shown to enhance average income both directly and indirectly.
The level of education does not significantly impact the statistical analysis of the dependent variable, as educational factors are often measured on a degree scale This lack of relevance may stem from many observations being recorded by household heads with only general education or from the reliance on experience over formal training in agricultural activities Both factors contribute to the notable difference when compared to household heads with specialized training.