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 calculated an index for land fragmentation and assessed 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 land, whether owned or rented (cited in Veljanoska, 2018).
Land fragmentation, as described by Van Dijk (2003), involves both the physical division of land and legal claims, representing two distinct layers This phenomenon can be assessed on various scales, which define what is considered "the whole." The research identifies four types of land fragmentation: ownership fragmentation, land use fragmentation, intra-farm fragmentation, and the separation of ownership from use While fragmentation may seem detrimental, it offers several advantages, including ecological and aesthetic benefits Additionally, it can mitigate the risk of crop loss due to disease or extreme weather, and certain agricultural practices require spatially separated parcels for diverse uses.
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 This phenomenon is particularly prevalent in developing nations, where fragmented land consists of multiple plots of varying quality Such diversity allows farmers to diversify their crops, manage labor demands effectively, and mitigate risks related to output and pricing.
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 has 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 noted by King and Burton (1982).
The average number of land parcels held by farmers in a specific region or country serves as a key indicator for assessing land fragmentation This evaluation is based on two primary characteristics: the average size of the holdings and the average parcel size, as noted by Edwards.
In 1961, a method was proposed to measure land fragmentation by assessing the percentage of a property owner's land area that is not contiguous with the farm Building on this concept, 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) measures land ownership distribution, where "n" represents the number of land parcels owned, "a" denotes the size of each parcel, and "A" is the total number of holdings An FI value of 1 signifies ownership of a single land parcel, whereas values approaching 0 indicate increased fragmentation of land holdings.
Dovring (1965) evaluated land fragmentation by calculating the distance a farmer must travel to reach each of their plots, returning to their farm after each visit (cited in Kadigi et al.).
The K-index, proposed by Januszewski in 1968, quantifies land fragmentation by calculating the ratio of the number of land parcels to the distribution of parcel sizes This metric is essential for understanding land use dynamics, as it accounts for the frequency of trips made to various property areas without necessitating a 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 This index has three key characteristics: first, fragmentation rises in proportion to the number of land parcels; second, it increases with a wider size range of smaller parcels; and third, fragmentation diminishes as 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 incorporates these factors.
The fragmentation index (P i) for 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 a notable limitation: 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) measures land concentration, where \( A_i \) represents the area of the \( i \)th parcel and \( A \) is the total area of the farm A value of zero indicates that the land is concentrated in a single parcel, suggesting that farmers possess only one piece of land Conversely, a value close to one indicates that families own 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 strategically 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 noted by various studies (Simons, 1985; Niroula and Thapa, 2005; Tran and Vu, Nguyen et al., 2020; Wang et al., 2021) 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 is irreplaceable (Mundlak, 2000).
The Cobb-Douglas production function and Stochastic Frontier Approach are commonly utilized in production theory to assess the impact of land fragmentation on output and production efficiency, as demonstrated in studies by Pham et al (2007), DiFalco et al (2010), Manjunatha et al (2013), Deininger et al (2017), and 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.
Research shows that economic theories and empirical evidence provide a strong basis for understanding how land fragmentation impacts 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) note that this fragmentation not only accelerates land 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.
Land fragmentation in China presents both challenges and opportunities for agricultural production He (2014) highlights that while land fragmentation negatively impacts production costs in some irrigation districts, it can also mitigate production risks Research indicates that smaller crop land sizes correlate with lower profitability, as small-scale farmers often rely on labor instead of mechanization, leading to reduced agricultural output (Markussen et al., 2016) Ciaian et al (2018) further note that smaller-scale production necessitates more labor, which can hinder the cultivation of crops that thrive on larger plots In India, fragmented land increases wheat cultivation costs, favoring labor over machinery, which may benefit self-sufficient households but not those targeting market sales (Deininger et al., 2017) Conversely, Blarel et al (1992) argue that land fragmentation can enhance risk management for farmers by diversifying farming patterns and alleviating labor shortages during harvests, as seen in Ghana and Rwanda.
Research has shown that land fragmentation, often presumed to be inefficient and detrimental to yield, may not significantly hinder land productivity Contrary to initial beliefs, the private benefits of fragmentation can equal or exceed its costs Policymakers are encouraged to understand the underlying causes of fragmentation, which stem from inefficiencies in property, labor, credit, and food markets Studies conducted in China indicate that land fragmentation can lead to substantial economic costs, while findings by Ali et al (2019)
Fragmented land serves as an effective strategy 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, despite encouraging crop diversity This aligns with earlier studies (Simons, 1985; Rembold, 2003; Pham et al., 2007), which indicate that fragmented land complicates cultivation, machinery use, and irrigation management Consequently, a notable 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, land fragmentation negatively impacts production outcomes, with effects varying based on the land usage index employed Fragmented land increases production organization costs, hinders innovation adoption, leads to harvest losses due to irregular plot shapes, and raises labor costs from dispersed land parcels Consequently, these factors contribute to substantial declines in productivity, revenue, profitability, and overall efficiency, indicating that fragmented land is detrimental to agricultural business performance.
2.3.2 The effect of land fragmentation on household income
Summary and conceptual framework
This chapter explores various theories and methods related to land fragmentation, highlighting its prevalence in some 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 certain conditions The agricultural environment of a specific location plays a crucial role in determining the effects of land fragmentation on farmers' interests To accurately assess the impact of land fragmentation on agricultural productivity, it is essential to validate findings through specific scenarios.
Household income serves as the expected dependent variable, enabling a direct assessment of land fragmentation's impact, particularly in Vietnam, where most rural households rely on agriculture for their income Evaluating the effects of land fragmentation on the income derived from agricultural activities is therefore a logical approach To analyze this relationship, a production function, such as the Cobb-Douglas function, will be employed with minimal changes in inputs Additionally, it is essential to incorporate various independent variables based on the data set used to comprehensively understand household income from agricultural pursuits in rural Vietnam.
This study aims to investigate the impact of fragmented landholdings on rural income, considering additional factors such as ethnicity, age, marital status, livestock activities, working-age population, household size, and types of land ownership The objective arises from conflicting findings in previous research, with some studies suggesting that land fragmentation negatively affects household income, while others propose that factors like work skills, education, and social networks have become more significant than land 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, with most of it 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 divided into four non-contiguous plots While large agricultural tracts constitute only about 10% of the total land area, they dominate the farmland landscape Although private corporations manage the majority of farmland, the state retains ownership and control over these businesses The primary aim of state-managed agricultural land is to produce resilient crops year after year, highlighting the significant issue of land fragmentation in Vietnam.
The OECD reports that land acquisition for expanding cattle production has commenced, yet the agricultural industry in Vietnam remains nascent, with only a small fraction of plots exceeding 2 hectares This fragmentation of land into smaller parcels poses a significant challenge to agricultural growth, which aims to meet the demands of an increasingly discerning consumer market Evidence indicates that agricultural output has notably declined due to the country's fragmented land, a consequence of rapid urbanization over recent decades.
(Pham et al., 2007; Tran and Vu, 2019; Markussen et al., 2016; Nguyen et al., 2020)
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 disorganization restricts the ability to develop consistent, large-scale production and limits access to essential resources such as transportation, irrigation, mechanization, and new technologies Consequently, valuable land that could be utilized for agricultural plots is wasted, discouraging both individuals and corporations from investing in the sector.
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 pieces 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, benefiting from significant policy support for land consolidation.
The Land Law (2013) aims to support investment activities by extending agricultural land use terms, reflecting significant changes in Vietnam's land policy This law promotes effective land use through measures that facilitate land accumulation and concentration for modern agricultural production Households can receive land allotments for up to 50 years, with extensions available for continued production needs However, expanding production often requires the relocation and integration of multiple units to form larger companies, posing challenges due to land use limits for investors Therefore, future land policies must be adjusted to introduce new regulations that enable legally binding transactions and foster 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 program 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 involves various factors, including the size and shape of individual plots, the distance between them, and travel time from home to the plots (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 was selected to evaluate land fragmentation, represented by the formula: 1 – (∑a_i^2).
The land fragmentation index ranges from zero to one, where a value closer to one indicates higher land fragmentation, meaning the household possesses multiple plots Conversely, a zero index signifies that the farm consists of a single plot, reflecting complete land consolidation.
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 (Debertin).
In 2012, various functions were developed from the Cobb-Douglas production function to address the issue of constant elasticity of input, with the exception of the Spillman function and polynomial forms Despite these alternatives, this study focuses on the Cobb-Douglas production function, as the parameters of the inputs remain unchanged.
The Cobb-Douglas Production Function serves as the theoretical foundation of this study, illustrating the relationship between input factors and output (E
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 the form of a double-log function (Ravallion & Van de Walle, 2008) This approach is appropriate as it accounts for various inputs, including age, marital status, and education level, in assessing 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), 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 on household access to productive resources, including land, physical, financial, human, and social capital It aims to uncover the reasons behind restricted resource access for certain households and how they acquire these resources Additionally, the questionnaire encompasses a broad range of topics, such as 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 survey 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 commonly 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 excluded the revenues and costs 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, focusing on soil-related factors such as fertility and irrigation, as well as land slope The research encompasses various aspects, including the general characteristics of households, agricultural land details, crop cultivation, land transactions, livestock, forestry, fisheries, market access, employment, and income sources Additionally, it considers extension services, household expenses, savings, credit, social capital, 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 consists of the Northern Highlands, represented by Lao Cai, Dien Bien, and Lai Chau; Region 3 covers the Central Highlands, with provinces such as 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 variances along the north-south and highland-lowland axes, with each region exhibiting notable disparities in the research findings.
The VARHS 2018 data on 15,442 acres of land was categorized by ownership type and land function This classification includes various land uses such as annual and perennial crop cultivation, forests, ponds, lakes, grasslands, 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 calculations exclude revenues and expenditures related to 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 was used for 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
Other 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:
The natural logarithm of per capita family household income from agriculture, denoted as LnY, is influenced by various household-level factors These factors, represented by the vector X, include household size, the number of laborers, livestock status, and characteristics of the household head such as age, education, marital status, and ethnicity.
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 must 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 employed to identify external instruments that either failed to meet requirements or were unavailable (Lewbel, 2012) The analysis revealed that the instrumental variables used to compute the land fragmentation index included the average slope of the parcel and variables from regions 1 and 2, which were chosen due to their representation of 71.75 percent of the surveyed households The average slope of land plots was categorized into four levels based on household size, utilizing data from the VARHS dataset Additionally, independent variables encompassed characteristics of the household head, such as marital status, ethnicity, education level, and age, as well as household-related factors like household size, number of laborers, and livestock production activities The education of household heads was classified into three groups: those with general education, those with short-term or professional diplomas, and those with college diplomas or higher Lastly, the livestock status variable served as a dummy indicator of whether a household participated in agricultural breeding activities.
To determine the reliability of the instruments used in this thesis, 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 unique households The survey reveals that 47.25 percent of these households belong to the Kinh ethnic group, while 52.75 percent are from other ethnic backgrounds Notably, 17.54 percent of respondents reported no marital status, with the majority of household heads (82.46 percent) being married The average age of household heads is approximately 52.7 years, and a significant 88.36 percent have not pursued education beyond primary school, with only 3.82 percent holding a college degree The number of residents per household varies from one to twelve, with an average of four to five individuals living in each surveyed home.
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 the land fragmentation index, agricultural income index, number of households, and household labor across various geographic regions in Vietnam The data indicates that northern provinces are facing considerable land fragmentation, while highland provinces exhibit lower household fragmentation rates Notably, Central highland provinces generate significantly higher agricultural income compared to other regions Additionally, households in the Northern Highlands are larger, averaging over five members, whereas other provinces maintain a more consistent household size Furthermore, while Northern lowlands, Central highlands, and Southern lowlands have similar household labor numbers, the Northern lowlands show a notably 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 UMB ER O F HO USE HO 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 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 attained Data collection reflects that 88.34 percent of respondents had not undergone intensive training, while only 175 heads of households had received short-term training Additionally, 11.64 percent of survey participants reported having completed short-term or medium-term training, or achieved education levels from intermediate to higher.
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 notably higher number of households within that specific workforce range.
Figure 6: Scatter of Land Fragmentation and Region
N UMB ER OF HOUS EHOL D
The scatter plot comparing the land fragmentation index across four regions shows consistent values ranging from 0 to 1 Notably, the southern regions (3 and 4) exhibit significantly less land fragmentation than the northern regions (1 and 2) This disparity may be attributed to the predominant village culture in the northern delta provinces, where cultural and historical factors lead to a higher land fragmentation index In these areas, it is common for parents to divide family farmland among their children as they reach adulthood, contributing to the fragmentation.
Figure 7: Scatter of Land fragmentation index and log of agriculture income per capita
The land fragmentation index significantly impacts the dependent variable in the research, as evidenced by a scatter plot illustrating the relationship between the land fragmentation index and the logarithm of average family income from agriculture 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 all households in these areas.
The "married" status of households constitutes a significant portion across all regions, indicating a correlation between marital status 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 UMB ER OF HOUS EHOL D
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 income per capita compared to those with fewer employees.
Figure 10: Scatter of age of household head and log of agriculture income per capita
The analysis reveals that the logarithm of annual agricultural income per person is distributed across the plot, correlated with the age of the individuals Notably, 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 from the dataset reveal that land fragmentation negatively impacts agricultural income per capita A significant negative correlation exists between agricultural income and factors such as the number of employees, livestock activities, and the ethnicity of the household head Additionally, instrumental variables, including the slope of the land at the household level and regional distinctions between the northern lowlands and 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, consistent with previous research on land fragmentation in Vietnam Additionally, the income disparity between Kinh-headed households and those led by individuals from other ethnic groups is projected to be 42.55% if all other factors are held constant, reflecting the characteristics of the data set collected in specific provinces.
The analysis reveals that ethnic minorities represent a significantly higher proportion in the region, with a deletion rate of 42.55 percent 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 other factors remain constant Furthermore, households led by married heads exhibit a 43.83 percent higher predicted per capita income compared to those with single heads, under the same conditions Engaging in livestock activities can lead to a per capita income that is 147% higher than that of households not involved in such activities, likely due to the higher selling prices of livestock-derived products Additionally, for every extra working-age individual in a household, there is an anticipated decrease of 12.6% in per capita income, while household size and the education level of the head do not significantly impact income calculations based on employed family members.
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 tests 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 research 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 practical experience over formal training in agricultural activities Both factors contribute to the notable difference when compared to household heads with specialized training.