Through using SMART PLS 4.0 software, the results demonstrate:
Measurement model test results (Fig 4)
An outer loading value of 0.7 or higher is significant when evaluating the quality of observed variables of the factors
(Hair et al., 2014). Through PLS-SEM algorithm analysis, the outer loading coefficients of the variables are shown from 0.538 to 0.889, of which three variables: LUE2 = 0.538, LUE4 = 0.593 and SF1 = 0.632 are in the range of 0.5 to less than 0.7 (Table 2).
The findings of Scale reliability of Cronbach’s Alpha, CR, AVE indicate that all factor structures have good reliability when Cronbach’s alpha reliability coefficient and Composite reliability coefficient (rho_c) are both greater than 0.7 and the AVE indexes are all 0.5 or higher, all factors ensure convergence and met the requirements (Table 3).
The Discriminant validity demonstrate that each variable’s AVE square root value is higher than the latent variables’ correlation coefficients and the concepts achieve discriminant value when the HTMT Index is less than 0.85 (Table 4).
Structual model testing results (Fig 5)
The model structure and hypotheses are illustrated in the structural equation model (SEM) and detailed in Table 1. Three paths lack statistical significance, as indicated by p-values exceeding 0.05: LP -> LUE at 0.078, LP -> LB at 0.396 and SF -> LB at 0.498. Conversely, the remaining nine hypotheses demonstrate support for causality, with p-values below 0.05. Specifically, the factors EF, Price and UF have a significant impact on LB; EF, GF and UF influence Price; LB images affect LUE; and LP mediates the relationships between Price and LB, as well as LB and LUE.
The LP has significantly influenced on both the relationship of Price - LB and LB - LUE with p values 0.045 - 0.000 respectively (Table 5).
Table 5 indicates that the Land price is significantly influenced by Economic and Urban factors with coefficients of 0.441 and 0.334, respectively. The urban also possesses an impact factor of 0. 324 on the Land bank. Land price and Economic factors are the second and third variables, with levels of 0.294 and 0.278, correspondingly.
The pricing system for land has begun to enhance urban land productivity in a positive and significant way
(Du et al., 2016). Creating a database contains all pieces of land price; Promote real estate values to calculate taxes, fees, other associated financial obligations. In Vinh Long context, to facilitate the creation of land banks in case land prices fluctuate, it is necessary to contribute capital in the form of land use rights to project implementation, harmonizing interests between the State and the investors, private sectors and land users.
The Land Bank has a significant influence on Land use efficiency, with an impact coefficient of 0.410. Geographic influence on land price is represented by a coefficient of 0.206 (Fig 5).
Land value is not only depended to the physical characteristics of a building but also it is depended to the built environment surrounds to that building. The streets were chosen as the research scale for measuring the spatial effects of the change process of land values in residential areas
(Kubat and Topcu, 2009). Geographic factors (GF) including land plots adjacent to the road, Land plot located at intersection, Size and shape of land parcel, Type of soil, Proximity to the urban center have essentially impacted on land price since it has the potential to be a highly profitable business. As a result, depending on the investor’s business goals, consider choosing a suitable location to implement projects in order to save costs and generate long-term returns. Local governments should classify industrial sectors based on geographical advantages to plan industrial land-use arrangement.
The planning allocation rate directly affects the decision to approve investment projects. The local authorities should take consideration of determining planning allocation targets is an important task that needs to be carefully investigated to best suit the reality of the method. If urban development policies fail to include precise location, area, boundary and spatial planning on the map, the likelihood of success is low. That means land allocation, land lease and land use purpose changes must adhere to approved annual land use planning and plans as outlined in the Land Law. Thus, there is a focus on creating and organizing land use plans that can be quickly adapted and are highly feasible.
The VIF quantifies the extent of multi collinearity among independent variables. As presented in Table 6, all VIF values are below 3, with the exception of the GD->QD value, which is 3.005. This indicates that the multi collinearity associated with this variable is not substantial, thereby having no significant impact on the testing of the research hypothesis and not constraining the R squared value. (Table 6).
Table 7 R- square illustrates that Price’s adjusted R-squared is 0.648, indicating that the independent variables GF, EF and UF account for 64.8% of the variation in the Price variable. With an adjusted R-squared of 0.235 for LUE, LB explains 23.5% of the variation in LUE. The R-squared value for LB is 0.518, signifying that the independent variables EF, UF and Price collectively explain 51.8% of the variation in the LB variable (Table 7).
In Table 8 of f-square, the Economic factor exerts a large impact on the land price variable (0.396), while the Urbanization factor demonstrates a medium impact (0.190). In relation to the Land Bank variable, the Urbanization, Economic and Land Price factors each exhibit a small impact, falling within the range of 0.02 £ f square £0.15, whereas the Social factor shows no impact (0.002)
(Cohen, 1988). Regarding the Land Use Efficiency variable, the Land Bank reflects an average impact level (0.171). The Legal Policy factor influences the relationship between Land Bank and land Use Efficiency with a medium impact (0.195) and has a small impact on the connection between Land Price and Land Bank (0.032) (Table 8).
The Q2 predictive relevance assessment involved employing the cross-validated Predictive (Q2) approach to evaluate the predictive accuracy of the structural model, as outlined by
Hair et al., (2014). This serves as a benchmark for determining the cross-validated predictive significance of the PLS path model. The Q2 index is regarded as a measure for assessing the overall quality of the component model. To derive this coefficient, the research will implement Blind folding analysis using Smart PLS, followed by an evaluation based on the specified criteria:
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.
Table 9 shows the values of the Q2 coefficient for the variables LUE, LB and Price are 0.115, 0.334 and 0.405, respectively, all exceeding the threshold of 0. The research model is of high quality and suitable, demonstrating a correlation between the dependent and independent variables (Table 9).
Hypothesis testing
The outcomes of the hypothesis test, as displayed in Table 10, all factors meet the reliability and validity criteria. Yet, the structural model’s significance in testing the study’s hypothesis is only confirmed nine hypotheses met the criteria while hypothesis H6 (with P values = 0.498 > 0.05) was not accepted. All hypotheses H1, 2, 3, 4, 5, 7, 8, 9 and 10 show a statistically significant association at the 5% significance level (Table 10).
LB acts as a go-between for Price - LUE, EF - LUE, UF - LUE and SF - LUE pairs, while LP is a moderate factor in the Price - LB and LB - LUE relationships (Fig 3).
The relationship between LP and Price on LB is demonstrated by a t-test p-value of 0.045, which is less than 0.05, indicating that Legal Policy has an impact on Land Bank. Therefore, the legal policy governs the connection between the price and land reserves. The recorded regression coefficient (O) of 0.076 indicates that altering the land policy will result in either an increase or decrease in land bank
via changes in land price.
The t-test p-value of the impact of legal policy on land use efficiency is less than 0.05, indicating that the relationship between LP (legal policy) and LB (land bank) affects LUE (land use effectiveness) significantly. Regression coefficient (O) of 0.266 > 0 indicates that modifications in land policy will affect land use efficiency, either positively or negatively,
via land bank.
