Some Policies on Agricultural Land Use Rights

Wald test:

With the hypothesis H 0 , all coefficients are not statistically significant. We choose the testing criteria:

𝑊𝑎𝑙𝑑 = 𝑛𝜔 2 𝛽 ̂ 𝑇[𝐷 𝑘 𝑘 ] − 1 𝛽 ̂ ~𝜒 2 (𝑘) (2.32)

𝜏 𝜏

Maybe you are interested!

• The impact of agricultural land use rights on agricultural production efficiency in Vietnam - Approach by econometric models - 19
• Regulations on Donation of Agricultural Land Use Rights
• Impact Channel From Agricultural Land Use Rights On Production Efficiency
• Regulations on Planning and Agricultural Land Use Plans
• Limitations and Causes of Limitations in State Management of Public Land and Agricultural Land in Cau Giay District

𝜏

In which: 𝑛 is the sample size, 𝛽 ̂ is the 𝑘 × 1 vector of estimated coefficients, 𝐷 𝑘𝑘 is the covariance matrix of 𝑘 variables to be tested, 𝜔 2 = 𝜏(1−𝜏)

𝑓(𝐹 −1 (𝜏)) 2

If 𝑊𝑎𝑙𝑑 > 𝜒 2(𝑘) then reject the null hypothesis H 0 .

Likelihood ratio (LR) test:

To test whether to shrink or expand the regression from the original model, we call 𝑉 ̂ (𝜏) the residual of the more constrained model and 𝑉(𝜏) the residual of the less constrained model.

Select the test standard as statistics:

𝐿𝑅 = 𝜔 − 1 (𝑉 ̂ (𝜏) − 𝑉(𝜏)) with 𝜔 2= 𝜏 (1−𝜏)

𝑓(𝐹 −1 (𝜏)) 2

(2.33)

If 𝐿𝑅 > 𝜒 2(𝑘) then the model is extended.

2.5. Non-parametric regression method

In regression methods, the parametric approach is often used with the assumption that the regression function has a predetermined functional form, such as parametric linear models, because this simplifies both the estimation procedure and the economic interpretation of the regression parameters. However, the assumption about the functional form of the regression may be inappropriate, and the economic conclusions and recommendations may be inaccurate. Non-parametric regression can overcome this drawback, because it can estimate the relationship between the dependent variable and the independent variables without giving a specific functional form.

Consider the regression model:

𝑦 𝑖𝑡 = 𝑚(𝑥 𝑖𝑡 ) + 𝑢 𝑖𝑡 (2.34)

where, 𝑦 𝑖𝑡 is the dependent variable; 𝑚(. ) is an unknown function; {𝑥 𝑖𝑡 } is the set of explanatory variables; random error 𝑢 𝑖𝑡 and time t.

The non-parametric regression method for estimating the function 𝑚(. ) in (2.34) is often approached and used as a local estimation. This method considers the function 𝑚 at each point 𝑥 = 𝑥 0 and at neighboring points 𝑥 0 , then there is

𝑚(𝑥) can be approximated by 𝑚(𝑥 0 ) or by the better approximation [𝑚(𝑥 0 ) + 𝑚′(𝑥 0 )(𝑥 − 𝑥 0 )] (by Taylor approximation) assuming that 𝑚(. ) is a smooth function. Thus, the estimation of the function

𝑚(𝑥) in the vicinity of point 𝑥 0 will be converted into estimates of 𝑚(𝑥 0 ) and 𝑚′(𝑥 0 ) from the observations obtained. And the estimation of function 𝑚(. ) on a certain interval will be done in the same way for each sub-interval. One of the modern and popular methods recently to estimate function m(.) is the Kernel function (can be called Kernel non-parametric regression).

Kernel Regression Function

The two commonly used methods in Kernel nonparametric regression are the local-constant method and the local linear method. The local-constant method was introduced simultaneously by Nadaray (1964) and Watson (1964), so it is also called the NW (Nadaraya - Watson) estimator. This method does not require assuming a specific form of the regression function 𝑚 , but a set of neighboring data points 𝑥 will contain some information about the value of

𝑚(𝑥) . Therefore, an intuitive estimator for the conditional mean function 𝑚(𝑥) is a local mean. In which, the new improved method is a weighted local mean.

Suppose 𝐾 is a real-valued function with weights assigned. The 𝐾 function is usually a density function that can be symmetric and is called the Kernel function. Let ℎ be the bandwidth (also known as the smoothing parameter), which is a non-negative number that controls the size of the neighborhood. The notation 𝐾 ℎ (. ) = 𝐾(./ℎ)/ℎ . According to Nadaraya - Watson (1964) the Kernel regression estimate is given by:

𝑚̂ ℎ (𝑥) =

𝑛

∑

𝑖=1

∑ 𝑛

𝐾 ℎ (𝑥 𝑖 −𝑥)𝑌 𝑖

𝐾 ℎ (𝑥 𝑖 −𝑥)

(2.35)

𝑖=1

where n is the total number of observations 𝑥 𝑖 in the neighborhood of point 𝑥 . The Kernel function is often used

2

used is the Gaussian Kernel 𝐾 ( 𝑡 ) = (√2𝜋) −1 exp(− 𝑡 ) . Where, 𝐾

(. ) is the density function

2 ℎ

The kernel is calculated with the division ℎ , or Epanechnikov Kernel 𝐾(𝑡) = 3 (1 − 𝑡 2 ) with

4

|𝑡| ≤ 1.

Under the usual conditions of the Kernel function, Nadanaya (1964) proved that the estimate (2.35) is a robust estimate of 𝑚(𝑥) . However, the estimate with this method is often biased at the boundary values ​​and in places with non-uniform distributions.

The local linear method proposed by Fan and Gijbels (1992) overcomes the above-mentioned bias problem of the NW method with the following idea:

In each neighborhood of a given observation point 𝑥 , instead of assuming 𝑦 to have a constant value as in the NW method, it is assumed that 𝑦 is linear in this neighborhood. After some mathematical transformations, the parameter estimation problem is reduced to an optimization problem as follows: at each point x, find the corresponding α, β to minimize the distance between the actual value and the estimated value, measured as follows:

𝑀𝑖𝑛 𝛼,𝛽 ∑ 𝑖=1−>𝑚 (𝑦 𝑖 − 𝛼 − 𝛽 ′ (𝑥 𝑖 − 𝑥)) 2 𝐾(𝑥 𝑖 , 𝑥, ℎ) (2.36)

where, 𝑛 is the total number of observations 𝑥 𝑖 in the neighborhood of point 𝑥 and the summation is performed on observations 𝑥 𝑖 such that ||𝑥 𝑖 − 𝑥|| < ℎ where ℎ is the bandwidth chosen during the estimation process.

In which, 𝐾(𝑥 𝑖 , 𝑥, ℎ) is defined as follows:

𝐾(𝑥 𝑖 , 𝑥, ℎ) = ∏ 𝐾 𝑖 (𝑥 𝑖𝑗 , 𝑥 𝑗 , ℎ 𝑗 )

𝑘

where 𝑘 is the dimension of vector 𝑋 𝑖 .

The Kernel function for the continuous variable 𝑥 𝑖 has the form:

𝐾 𝑗 (𝑥 𝑖𝑗 , 𝑥 𝑗 , ℎ 𝑗 ) = 𝑘 𝑗 (

𝑥 𝑖𝑗 − 𝑥 𝑗

)

ℎ 𝑗

The thesis uses the Epanechnikov Kernel function, which is a function type that is quite commonly introduced in many documents. Specifically:

3

𝑘 𝑗 (𝑧) = { 4√5

(1 − 1 𝑧 2 ) 𝑛ê ́ 𝑢 | 𝑧 | ≤ √5)

5

(2.37)

0 customers were charged

The Kernel function for discrete variable 𝑋 𝑖 has the form:

𝐾 (𝑥

1 𝑛𝑢 𝑥 𝑖𝑗 = 𝑥 𝑗

, 𝑥 , ℎ ) = {

(2.38)

𝑗𝑖𝑗 𝑗

Wife of a man

If the split is too large, there will be many observations in the estimate of (2.36), the estimated variance is small but the bias is large. If the split is too small, there will be only a few observations for the estimate, then the variance of the estimate is large and the bias is small. To balance the trade-off between bias and variability, the thesis uses the cross-validation method proposed by Li and Racine (2004) in choosing the split for each variable.

The advantage of the non-parametric method is that it does not require a specific functional form for the correlation relationship as well as assumptions about the distribution of random errors. However, when the model has many independent variables, the estimation becomes complicated (Horowitz, 1993), due to the continuous fragmentation of data during the estimation process. Therefore, the thesis uses both types of methods to check the robustness of the results.

Chapter 3

CURRENT STATE OF AGRICULTURAL PRODUCTION AND

AGRICULTURAL LAND USE RIGHTS IN VIETNAM FOR THE PERIOD 2012 - 2018

Chapter 3 will discuss the history of agricultural land use rights development in Vietnam and the efficiency of agricultural production over time. At the same time, this chapter also provides statistical analysis of land use rights, sociological factors and agricultural production efficiency in Vietnam in the period 2012 - 2018 along with the correlation between them.

3.1. Some policies on agricultural land use rights

3.1.1. History of development of agricultural land use rights in Vietnam

Cultivated land in Vietnam has undergone various forms of ownership, depending on the respective social regime. During the feudal period, land belonged to the King and the King granted land use rights to court officials. Land was then distributed to farming households based on the number of male workers and could be redistributed after a period of three to five years according to population changes. After the August Revolution, in the North, land was reclaimed and redistributed. Cultivated land was confiscated from landlords and then transferred to peasants. During this period, 810,000 hectares (37%) of cultivated land in the North was distributed to 2.1 million households, an average of 0.4 hectares per household. However, land was then collectivized in the late 1950s (Pingali and Xuan, 1992).

After the reunification of the country in 1975, the Vietnamese government redistributed land in the spirit of state ownership by the entire people. By 1981, from Directive 100 of the Party, the government allocated land to farmers to cultivate and sell products to cooperatives under contracts at fixed prices, and surplus output could be retained for consumption or sold to private traders, while cooperatives continued to provide inputs and means of production, and farmers were responsible for managing crops and livestock on their land. In 1988, Resolution 10 came into effect, granting land use rights to households. Land was allocated to households in a relatively equitable process, usually based on household size. Land use terms were approximately 10-15 years for annual crops and one or two cycles of planting forests and perennial crops. Families retained control over farming decisions and production output. After this reform, overall productivity in rice production increased significantly (Pingali and Xuan, 1992). To increase

To strengthen tenure security, the government passed a land law in 1988, allocating land to households for 10 to 20 years on a renewable lease basis. As a result, the poorest households have become more affluent in their consumption, and there is no evidence of systematic malpractice by local authorities (Ravallion and van de Walle, 2004). The transfer of agricultural land use rights from collectives to individual households in 1988 was a key element of the “Doi Moi” process. However, the 1988 Land Law did not consider land use rights as property rights, and land users were not allowed to transfer their land use rights.

With the 1993 Land Law, household land use rights were further strengthened, allowing for the transfer, conversion, lease, mortgage, and inheritance of land use certificates (National Assembly of Vietnam, 1993). This was the third breakthrough in land issues in Vietnam. Land use certificates were issued to households for a term of 20 years for annual crop land and 50 years for perennial crop land, and could usually be renewed upon expiration. By 2000, nearly 11 million land use certificates had been issued to rural households, one of the largest rural land use rights programs in developing countries (Do and Iyer, 2008).

The 2003 Land Law continued to improve some conditions on agricultural land use rights for people. Specifically, the land use limit was abolished, the state did not limit the land use area but only limited the allocated land area (land allocation limit). Land use rights were recognized as a special type of property right, allowed to participate in transactions on the market through transfer, lease, conversion... Land was recognized to have value and the state determined the land price for management. The 2003 Land Law was promulgated not only to supplement the right to donate and give away land use rights but also to give land users the right to choose the form of land use. In the 1993 Land Law, the issuance of land use right certificates was managed by provincial agencies, which led to many delays and impracticality in land management. The 2003 Law improved this factor by stipulating that all authority to decide on land and land management for households, individuals, and residential communities falls under the authority of the People's Committee at the district level and all authority to decide on land and land management for organizations, religious establishments, and foreign investors falls under the authority of the People's Committee at the provincial level. In addition, the 2003 Land Law increased gender equality by including the "gender" factor in the law. The law recognizes and protects the equal rights between husband and wife in land use rights. The State has policies to create conditions for those directly engaged in agricultural production to have land for production, and to divide land in a widely dispersed direction to diversify

production. However, the Law still has some limitations, such as regulations on the duration and limits of agricultural land use, which are the cause of difficulties in accumulating land for investment in developing large-scale agricultural commodity production.

The 2013 Law has made significant changes in agricultural land use rights compared to the 2003 Law. For example, the land use term has been extended from 20 years to 50 years. This is a positive change, allowing people to invest long-term and change the crop structure. Households and individuals are granted 02-03 hectares of annual crops and 10-30 hectares of perennial crops without paying land tax. In addition, the 2013 Law provides more rights for agricultural land users so that they can convert, transfer, lease, sublease, inherit, and mortgage their land use rights. Specifically, the limit for individuals and households to receive transfers of agricultural land use rights has increased from 3 to 5 times (depending on the type of land and the region), in which households can transfer land use rights with an area no larger than 10 times the limit. This opens up the possibility of land accumulation to improve farming skills and economic efficiency in agricultural production. In addition, the State has also issued additional Decrees and Circulars to supplement and guide the implementation of the Law, specifically: Government Decrees 43 and 47 of 2014, Decree 01 of 2017, Decree 148 of 2020 ... on guiding the implementation of the Land Law, which supplements regulations on land prices and conversion of land use purposes. Most recently, Circular 09 of 2021 of the Ministry of Natural Resources and Environment has provided more specific instructions on the implementation of the Law, allowing the flexible implementation of the mechanism for converting land use purposes within the agricultural land group to better suit the market mechanism; this also creates more favorable conditions for land concentration for commodity production according to the farm economic model. At the same time, the Law also stipulates the regime of land use for farm economy, in which the State encourages the form of farm economy of households and individuals to effectively exploit land to develop production, expand scale and improve the efficiency of land use in agricultural production.

Despite improvements compared to the 2003 Law and the 2009 Amendment and Supplement Law, the 2013 Land Law still has some shortcomings. Firstly, according to the provisions of the 2013 Land Law, the term of use of agricultural land groups in general has been increased to 50 years, however, Article 37 still maintains the 10-year planning period. This shows that, although the land use term has been increased, in Vietnam today, the majority of agricultural land is still not guaranteed to be sustainable because changes in planning will change the form of land use. Secondly, although the limit on land use rights for households has increased, it is still limited to about 2 - 3 hectares, if

Concentrated land more than 10 times will be transferred to the State for lease; households and individuals who do not directly produce agriculture are not allowed to receive transfers or gifts of rice-growing land. Difficulties in agreeing on land prices when negotiating land transfer or lease make it difficult to negotiate land use rights. The renovation and construction of rice fields require high costs due to fragmented land; many embankments and plots, formed long ago, while the funding source for implementing planning items and renovating the rice field system is still lacking. Therefore, to concentrate land to expand production scale, investors will face difficulties in finding land sources as well as legal procedures. Thus, although the 2013 Land Law has innovated and expanded land use rights for individuals and households, the planning vision has not kept up with the growth rate of the economy, easily leading to instability in land use and exploitation, including agricultural land.

3.1.2. Vietnamese agriculture through the ages

Since the reform of the agricultural management mechanism in the spirit of Resolution 10 of the Politburo of the Central Executive Committee of the Communist Party of Vietnam (April 1988) and the implementation of the one-price policy (1989), the agricultural economy has had a remarkable growth. The average increase in food output in terms of paddy per year in the period 1986 - 1990 was 0.67 million tons, in the period 1991 - 1995 was 1.2 million tons, in the period 1996 - 2000 was 1.2 million tons. In general, in the first 15 years of reform (1986 - 2000), our country's rice output increased by 100%, the area increased by 28% and the productivity increased by 50%. The average growth rate of the respective periods was 3.38%; 5.06%; 5.0% (GSO, 2000). The structure of cultivated area has been gradually diversified. From 1986 to 1999, the proportion of food crops decreased from 91.2% to 72.9%, the proportion of non-food crops increased, especially industrial crops. In the period 2002 - 2012, the average income growth rate in rural areas was 19.1%/year compared to 17.5%/year in urban areas. In 1999, per capita income in urban areas was 2.3 times higher than per capita income in rural areas; This index decreased to approximately 2 times in 2012. The average income per capita of rural households increased from 225 thousand VND/month in 1999 to 275 thousand VND/person/month in 2002 to about 1.6 million VND/person/month in 2012. Along with the economic development process, the proportion of agricultural households gradually decreased from 79.58% in 1994 to 77.35% in 2001 and to 57.79% in 2011. Currently, household economy contributes 90% to the development of important agricultural sectors of the country such as rice, coffee, livestock...