Evaluating the Land Use Implementation Results Through Land Policy in Vinh Long Province, Vietnam

As discussed by Veršinskas (2022), Using land bank creation tools, analyzes and identifies European good practices on land banking. Specifically, research on many models and methods of creating land banks in Denmark, France, Germany, the Netherlands and Spain are based on legal tools, institutions and monetary funds to develop land and propose a policy. To create a land bank for the remaining countries in Eastern Europe and Central Asia, depending on the host country’s state structure and political institutions.
       
According to Marošan et al., (2014), based on institutional theory: the institutional framework changes according to the development of the land management system, administrative framework, as well as the goals to be achieved through the land bank. There are three types of rents, including differential rent of one from differential advantage in production; land rent differential two from the rent difference between unimproved land and land improved through capital investment; absolute rent from the scarcity of land; and monopoly rent from the existence of pure monopolies over scarce products.
       
Chuong et al., (2020) pointed out the theoretical basis for the mechanism to create a clean land bank, the system of current legal regulations of the government and Can Tho City on creating a land bank to attract investment, investing in Can Tho City, stating difficulties and limitations and proposing specific solutions. Other localities, such as Quang Binh and Gia Lai, have many authors who have proposed many groups of factors, including Policy factors, financial factors, planning factors, nature, infrastructure and other factors that affect land banks for economic and social development.
       
In addition, according to Park (2014), land prices and land taxes are formed based on regular analysis of profits from land use in the market mechanism to determine the differential land rent belonging to the State correctly and the land rent difference between two belonging to the investor. Vinh Long province is located in the center of the Mekong Delta (MRD), with a total geographical area of 1,525.73 km² (152,573 hectares). The population as of 1,029,015 in 2019 and the population density is relatively high at 674 people/km² (Vinh Long People’s Committee, 2019). Vinh Long’s economy has grown relatively high in recent years; by 2022, the economic growth rate have reached 11.28% and ranking 10^th in the country and 3^rd in Mekong Delta region (Vinh Long People’s Committee, 2023).
       
To continue the economic and social development of the Province, the People’s Committee of Vinh Long Province has implemented many land bank projects to improve infrastructure, serving to attract businesses and housing. The Province has issued many land policies, site clearance, compensation, support and resettlement in road expansion projects, urban expansion, etc, to attract companies and businesses interested in urban beautification. However, the land area is limited, in the total area structure of 152,573 hectares, including 03 main land groups as follows: Agricultural land group is 119,570 hectares, accounting for 78.37%; Non-agricultural land group is 32,986 hectares, accounting for 21.62%; The unused land group is 17 hectares, accounting for 0.01% of the total natural area (Vinh Long People’s Committee, 2023).
       
To have a land bank to carry out projects, the State creates a land bank through 02 mechanisms: Mechanism 1 - compulsory mechanism, forced purchase, that is, the State recovers the land from an owner to allocate land for other users to implement projects for investors to build public works - public investment projects; Mechanism 2 - voluntary mechanism, which means investors agree to receive land transfer from land users or participate in auctions of land use rights to serve production and business projects for private benefits. - private investment projects. However, following these two mechanisms, the State and investors are encountering difficulties, causing loss of costs, time, waste of land and ineffective land use.
       
According to Mechanism 1, the State must recover land to implement the project (National Assembly, 2013), which creates complaints from landowners that must be resolved for a long time. 2013-2020, the country received about 324,241 land complaints (MONRE, 2022). From 2013 to 2020, there were 1,375 land complaints (DONRE, 2021). Co Chien - Vinh Long Industrial Line, with a scale of 46 hectares, started to be deployed in 2004; 35 hectares have been assigned and leased to 37 organizations and individuals to use; the remaining 11 hectares have not been cleared for about 20 years yet-relocated households to hand over to investors cause of the complaint of land owners (Vinh Long People’s Committee, 2019).
       
According to Mechanism 2, a voluntary mechanism, investors agree with land users to buy, rent, receive capital contributions, or participate in auctions of land use rights to implement the project (National Assembly, 2013). However, the most prominent difficulty investors encounter is that the land owner does not agree to sell or lease the land, or the price is very high and the parties cannot agree.
       
Based on the current situation of the local land bank and the inadequacies of the current mechanism, research is necessary to find and identify factors affecting the land bank, such as planning laws, prices, the natural location of the land plot, etc., thereby proposing policy implications for land, construction and investment that impact better land use efficiency for Vinh Long province.

The study was conducted from 2020 to 2023 in the Vinh Long province included the following contents.
 
Data collection
 
Secondary data source
 
Using data from published domestic and foreign research (data from national and foreign magazines), statistical reports of state management agencies about resources and environment, surveys and data analysis, students can draw out factors affecting the land bank.
 
Primary data source
 
Random sampling was done by sending survey questionnaires to 170 officials, civil servants, lecturers in the natural resources and environment sector and people’s committees using the Google Forms tool.
       
According to Nguyen (2014), the minimum sample size is n ≥ 50 + 8*p, where p is the number of independent variables. So the minimum sample size is n ≥ 8*5 + 50 <=> n ≥ 90. The study surveyed 170 people, including all civil servants, public employees and land researchers.
       
Use a 5-level Likert scale with the following corresponding levels: 1 = completely disagree; 2 = disagree; 3 = neutral; 4 = agree; and 5 = strongly agree.
       
After adjusting to suit local practice, the scale measuring land bank factors results include four hidden factors with 34 observed variables used as the primary scale for the study. Four hidden factors (cause scale) that impact the Land bank factor include (1) Legal and policy factors, (2) Financial factors, (3) Planning and planning factors, (4) (Geographical) Natural factors of the land plot and the factor “Land bank” will impact the variable “Land use efficiency.”
       
The summary of the scale, observed variables and variable symbols are as follows:
 
Land bank variables (Decision)
 
 Land recovery (Decision 1), Receive transfer (Decision 2), Encroachment on rivers, canals, mudflats (Decision 3), Voluntary return of land (Decision 4), State recovery of land due to illegal violations Land law (Decision 5).
 
Variable land use efficiency (HQ)
 
Quick land allocation and land lease procedures (HQ1), Land area has been cleared (HQ2), Implementation progress and project coming into operation on time (HQ3), Bringing socio-economic benefits (HQ4), project has exploited land use (HQ5).
 
Variable law and policy factors (PL)
 
Regulations on land allocation, land lease, bidding for projects using land (PL1), Legal regulations on investment licensing (PL2), Legal regulations Law on construction (PL3), Tax incentives and investment incentives (PL4), Publicity and transparency of land information (PL5), 16. Regulations on compensation, support and resettlement (PL6).
 
Land finance factor variable (TC)
 
Land price (TC1), Investment capital from the state budget (TC2), Loan capital from credit institutions (TC3), Capital contribution by the value of land use rights (TC4), Aid from domestic and foreign organizations (TC5), Land Development Fund (TC6).
 
Planning element variables (QH)
 
socio-economic master plan (QH1), Land distribution put into use (QH2), Current land use status (QH3), Land use allocation criteria approved (QH4), urban and rural development planning (QH5), Sector planning (QH6) and Preparation and approval of land use plans (QH7).
       
The natural factors of the land plot (TN) are the following: Location (TN1), Area (TN2), purpose of use (TN3), Profitability (TN4) and Soil (TN5).
 
Data analysis
 
To evaluate the measurement model on SMART PLS, we evaluate on the following criteria:
-   Quality of observed factors variables: Outer loading has a good meaning of 0.7 or higher. Reliability Cronbach’s alpha scale reliability ≥ 0.7 (DeVellis, 2012), Composite reliability ≥ 0.7 (Hair et al., 2014).
-   AVE convergence (convergence) ≥ 0.5 (Hock and Ringle, 2010).
-   Discriminant Validity: Equal to HTMT table ≤ 0.9
   
After evaluating the measurement model on SMARTPLS, we continue to evaluate the structural model (SEM) based on considering the following factors:
-   Calculate the collinearity of independent variables with inner VIF < 5
-   Meaning of impact relationships in the model (Path Coefficients): P value < 0.05 is a statistically significant relationship and P value > 0.05 is statistically insignificant.
   
The original sample path coefficient has a positive sign for a positive effect (+) and a negative sign for a negative effect (-).
-   Evaluate the coefficient of determination R square. R-squared approaching 1 means the level of explanation for the dependent variable is high and approaching 0 means the level of explanation for the dependent variable is low. (There is no R-squared threshold for how much is passed or what is not) (Hair et al., 2017).
-   Evaluate the impact coefficient f square (f square) f square < 0.02: the impact level is minimal or has no impact. 0.02 ≤ f square < 0.15: trim impact level. f square < 0.35: average impact level. f square ≥ 0.35: significant impact level (Cohen, 1988).

Measurement model test results
 
The results shows that all factors structures are reliable when Cronbach’s alpha and Composite reliability coefficients (rho c) are more significant than 0.7 and the AVE indexes are all from 0.5. All factors ensure convergence (Table 1).
       

 
The analysis results shows that the square root value of the AVE of each variable is greater than the correlation coefficients between the latent variables and the HTMT Index is less than 0.9, so the concepts achieve discriminant value (Table 2).
 

       
Evaluate the quality of observed variables of the factors through PLS-SEM algorithm analysis (Fig 1), showing the results of the outer loading coefficients of the variables from 0.331 to 0.948, in which HQ5 = 0.331, QH7 = 0.635 and TN4 = 0.622 < 0.7 has less statistical significance, so these three variables are removed from the diagram (Table 3).
 

 

 
Structural model testing results
 
Table 4 shows that the VIF coefficients are all less than 5, showing no multicollinearity phenomenon between the independent and dependent variables.
 

       
Path Coefficients represent the P value of Natural -> Land bank = Natural -> Efficiency = 0.718 > 0.05, so it is not statistically significant (Table 5).
 

       
Except for the Natural variable, all the impact coefficients of the independent variables in the Original Sample column have positive signs. Hence, the impact relationships in the model are all positive. The order of impact from strong to weak on the Efficiency variable is QD (0.769) > PL (0.609). The order of impact from strong to weak on the land Bank variable is PL (0.581) > QH (0.290) > TC (0.204) (Table 5).
 
Test the coefficient of determination R²
 
The R² value of efficiency is 0.661, indicating moderate accuracy, meaning that land Bank factors predict 66.1% efficiency. The land Bank’s R² value of 0.535 shows a moderate level of accuracy, meaning that the Efficiency factor predicts 53.5% of the land Bank.
       
Results from the Table 6 show that independent efficiency variables include the land bank variable, which has a strong impact (1.131) and the law and land bank variables, which have a negligible impact (0.036).
 

 
Independent variables of the land Bank
 
Law variables have a substantial impact (0.712), planning and finance variables have a negligible impact (0.084, 0.048) and natural variables have no impact.
 
Test the relevance of forecast Q² 
 
The cross-validated prediction (Q²) method was used to measure the predictive fit of the structural model (Hair et al., 2014). It is the criterion for evaluating the cross-validated predictive significance of the PLS path model (Fig 2). The Q² index is considered an index to evaluate the overall quality of the component model.
 

       
The study uses blind folding analysis in Smart PLS to estimate the coefficient. Then, evaluate based on the following levels:
·  0 < Q² < 0.25: low level of forecasting accuracy.
·  0.25 ≤ Q² ≤ 0.5: average level of forecast accuracy.
·  Q² > 0.5: High level of forecast accuracy.
       
The Q² coefficient is a standard to determine the model’s predictive ability. The Q² coefficient values of the variables efficiency and land bank are 0.340 and 0.509, respectively, more significant than 0. It shows that the research model is good quality and appropriate, with a correlation between the secondary and independent variables.
 
Hypothesis testing
 
The results of the hypothesis test in Table 7 shows that all factors satisfy the criteria of reliability and validity. However, the structural model used to test the study’s hypothesis is only statistically significant when p-value ≤ 0.05. Five hypothesis were accepted based on the criteria for testing the structural model and 01 hypothesis H4 (P values = 0.718 > 0.05) was rejected. All hypothesis H1, 2, 3, 5 and 6 have a statistically significant relationship at the 5% level. However, hypothesis H4 is not statistically significant.
 

       
In which QD plays an intermediary role between the pairs of variables PL - HQ, QH - HQ, TC - HQ, TN - HQ and the variables PL, QH, TC and TN have the same statistically significant impact on the variable. HQ.
       
The t-test p-value of the relationship Law × Land bank impacts on efficiency is 0.026 < 0.05, showing that the product of Law × Land bank impacts efficiency. Thus, the law regulates the relationship between the Land bank and Land use efficiency. Original sample regression coefficient (O) = 0.124 > 0 shows that changes in the law will increase or decrease land use efficiency through the Land bank.
       
Table 5 shows that the law substantially impacts the Land bank (original sample: 0.581). It proves that the state managed land by law and strict regulations from land allocation and land lease, bidding for project implementation to investment licensing, construction and taxation. Therefore, the Government of Vinh Long province must promote and effectively implement land law guidelines and policies on creating land banks, ensuring investment attraction through reasonable support capital and conditions to have clean premises for investment.
       
Next is the planning factor, which has an impact coefficient of 0.290 on the land Bank. The planning process is carried out according to the principle of planning from general to detail; Lower level planning must be consistent with the upper-level planning and the higher level’s planning must reflect the needs of the lower level. In addition, the master plan for socio-economic development is an activity aimed at concretizing the socio-economic development strategy to determine the appropriate spatial sector structure to help socio-economic development. Sustainable society also has a guiding role for lower-level plans. It is a factor that directly impacts land use planning and land bank development, demonstrating the significant role of the master plan for socio-economic development in land bank development. It also shows that there needs to be research on the overall socio-economic development plan in implementing land bank development work, creating a two-way relationship to ensure the land bank’s development work is effectively deployed.
       
Planning norms affect the decision to approve investment projects. Projects whose scale exceeds planning norms are considered inconsistent with the planning and must be implemented. Therefore, determining planning norms is an important task that needs to be carefully investigated to suit the reality of the area best. Organize the management of land use purposes from the time of planning and announcement. Pay attention to calculating land use efficiency when formulating local land use planning projects, thereby allocating appropriate land use plans without waste. Strictly manage the implementation of detailed land use and construction planning; Resolutely handle cases of non-compliance with planning.
 
Finance is the final impact factor, with an impact coefficient of 0.204, showing that capital sources from the budget, credit institutions and foreign aid organizations are often difficult to access and the local state does not have an initiative that depends on the factors of the account holder. Therefore, competent authorities must pay attention to investment attraction policies, especially investment projects that use land. At the same time, there is a mechanism to encourage investors to advance capital for land bank creation, encouraging projects used for production and business purposes to invest in industrial parks and industrial clusters in the area.

The research results have identified three factors: Law, planning and finance that affect land banks, which are intermediate factors directly impacting land use efficiency in the Vinh Long area. Analysis results show that 65.5% of the variation in land use efficiency can be explained by the variation of 4 factors (Land bank, Law, Planning and Finance). Land bank is the most substantial direct impact factor, with an impact coefficient of 1.131. Land bank is the most critical factor among the six factors of the analytical model. The following concerns the Legal factor affecting the Land bank with a coefficient of 0.712. The remaining factors related to Planning, Finance and Law to efficiency have coefficients of 0.084, 0.048 and 0.037 respectively. Land bank and legal factors are essential for land use in Vinh Long area. The remaining 34.5% is explained by factors other than the model.
       
The research has made several significant contributions to perfecting theories and methods of state land management, providing solutions to improve the effectiveness of state management in this field in the Vinh Long. From the suggestions provided by the research, public managers can learn to predict and plan effective land use development strategies suitable to local economic and social conditions.
       
Besides the results achieved, the above study has limitations, such as the convenient sampling method, so the research sample needs to be representative. Therefore, the authors propose using a random quantity collection method in the following research direction. Besides, the new topic only focuses on organizations and individuals performing state management tasks and scientific research on land in Vinh Long area. The following new research direction may be expand the survey scale of non-state corporations and companies in other locations to compare research issues across regions.

The authors have no conflicts of interest to declare. All co-authors have seen and agree. We certify that the submission is original work and is not under review at any other publication.