Estimating farmers’ willingness to pay for photovoltaic industry to improve agricultural green resources and environment

Introduction

Agricultural green development is based on high-quality resources and the environment (Riti et al., 2021). In recent years, various countries have adopted a series of policy measures to promote the improvement of agricultural resources and the environment (Brock and Reinhilde, 2014). On one hand, they promote the development of the green energy industry and use external forces to help the green development of the agricultural modernization industry and the upgrade of green production technology, and on the other hand, they improve the utilization rate of agricultural waste resources and use the internal power of agricultural production to improve the green circulation ability of agriculture (Li et al., 2022). As a kind of clean energy, solar energy is not only an alternative to power generation but also integrated with farm production and management in various countries based on local climate and land reserve and produced good benefits (Nizam et al., 2013). It can be seen that the development of the photovoltaic industry can drive the upgrading and transformation of agricultural green production technology, improve the utilization rate of agricultural green resources, drive new economic benefits, and play a significant role in promoting agricultural green development. However, the quality of the agricultural production environment is still not optimistic, and an imbalance and insufficient development of agricultural green resources and the environment still exist. The development of the agricultural green energy industry has just started, and the radiation improvement of agricultural green resources and the environment still needs to be continuously performed (Lu et al., 2020; Yu, 2020; Rahim, 2022). To sum up, in the important development stage of the agricultural green energy industry, how to investigate the effect of the green energy industry on the improvement of agricultural resources and the environment from the perspective of truly benefiting rural residents is of great significance not only for farmers to improve their awareness and utilization of green energy technology but also for farmers to upgrade production technology and optimize the agricultural production environment.

Agricultural green resources and the environment have typical public goods’ attributes, which can bring beneficial effects to other economic parties and have positive externalities. When the development of green energy industries has a certain incentive effect on the improvement of local agricultural green resources and the environment, beneficiaries are willing to pay for the resources of potential benefits of environmental public goods (Carson et al., 1992), which involves the preferences of beneficiaries and is related to the non-use satisfaction benefits of environmental public goods (Ressurreio et al., 2012). Taking the photovoltaic industry in the green energy industry as an example, this study measures the value of the beneficiaries to the photovoltaic industry to improve agricultural production resources and the environment and maintain the current state without declining so as to make these resources and environment sustainable existence (existence value), use by future generations (bequest value), and their future use (option value) (Fitzpatrick et al., 2017). It is also the content of farmers’ willingness to pay (WTP) the non-use value of agricultural resources and environmental improvement.

The contingent valuation method (CVM) is the most important non-use valuation method for public goods (Waldo et al., 2016). It is widely used in the non-use value assessment of resources and the environment, such as environmental economy, solid waste treatment, ecosystem services, and biodiversity protection (Tonin, 2019). The CVM model application of the theory of consumer surplus and welfare changes according to the state of resources, and the environmental impact assessment method used to estimate the value of public goods, namely, by constructing illusions in the market, and evaluate people’s WTP of resource environmental quality will improve to a certain extent or a certain degree of deterioration of intention as the foundation, the resources, and the environment, such as currency valuation of non-market goods or services (Ferreira and Marques, 2015). It plays an important role in the cost-benefit and damage analyses of resources and environmental public policies, as well as the formulation, implementation, and effect evaluation of related policies (Hanley and Czajkowski, 2019).

When the CVM is applied to the evaluation of familiar public goods dominated by the use value, construct validity estimation can be generated (Ronald and Glenn, 1995). However, when the CVM is applied to the evaluation of public goods dominated by the non-use value, the validity and reliability of the results have generated many debates (Baron, 2017). Embedding effects are the main internal debate (Bateman et al., 2004) and one of the main sources of error (Veisten et al., 2004) when researchers use the CVM. It is the most difficult one to deal with among the main deviations of the CVM (Hausman, 2012). The embedding effect refers to the fact that the same public goods are valued at a lower level as part of more inclusive public goods than they are individually (Kahneman and Knetsch, 1992), or that there is little difference in valuation between the public goods and the more inclusive public goods (Ronald and Glenn, 1995), or that different investigations of the same public goods result in widespread changes in the WTP (Hausman Hausman, 1994). The respondents usually have no experience in the selection of public goods to be estimated, so they are more likely to be affected by the embedding effect in non-use value assessment (Schulze et al., 1998). The National Oceanic and Atmospheric Administration (NOAA) recommends internal consistency tests be performed to assess the validity and reliability of CVM studies (Kerry et al., 1996). The scope test is a standard method to examine the existence of an embedding effect (Arrow, 1993).

At present, the research results on embedding effects mainly focus on explaining or verifying the phenomenon of embedding effects from the aspects of diminishing marginal utility (Pinto et al., 2016), survey design or execution ability defect (Ojea and Loureiro, 2010), substitution effect, income effect (Jorgensen et al., 2013), etc. From the existing typical cases of embedding effect research, there are many achievements in selecting rivers, wetlands, natural parks, and forests as research objects (Grammatikopoulou and Olsen, 2013; Giguere et al., 2020). At present, existing studies on the WTP for agricultural resources and the environment based on the CVM mainly focus on rural tourism, cultural industry, forestry, and other aspects (Tonin, 2019), while the research on the WTP for the improvement of agricultural resources and environment based on the green energy industry is obviously lagging behind. Moreover, no research result is found on the embedding effects of choosing the willingness of the green energy industry to pay for the improvement of agricultural resources and the environment as the research object.

In general, an area where a green energy industry is located will not only promote the improvement of local agricultural green resources and the environment but will also benefit its nearby areas. The radiation degree decreases with the length of this distance. When the improvement of agricultural resources and the environment in the region where the green energy industry is located is valued as a component of the improvement of resources and the environment in a larger region, due to factors such as saturation degree, diminishing marginal utility, or substitution effect, the embedded-effect phenomenon occurs such that the valuation result is lower than the individual valuation result or the sum of the independent valuation results exceeds the overall valuation result.

Therefore, a reasonable analysis and test of farmers’ WTP for the photovoltaic industry to improve agriculture and green production resources and the environment and the embedding effect can better fill the gap in the existing literature and provide a reference value for relevant empirical research. This study chooses Ningxia Province, a typical representative province of the photovoltaic industry in China, as the study case. Using the double-bound dichotomous CVM model (Changlin et al., 2016), which can reduce the WTP bias caused by market distortions, we used a C-optimal design to revise the bidding value of the payment scheme (Kim, 2009), combined the double-bound dichotomous CVM with embeddedness effect research, and conducted a study on farmers’ WTP for the photovoltaic industry to improve agricultural green resources and the environment. At the same time, analysis of the substitution effect, convergence effect, and influencing factors in different regions, in order to promote the development of the photovoltaic industry in different regions according to local conditions, effectively improves the utilization rate of the photovoltaic technology and increases the radiation benefits of a new energy industry to agricultural green production and environmental improvement. It is expected to provide an empirical and theoretical basis for cost-benefit analysis of agricultural green resources. It also provides an empirical reference for the compensation and subsidy measures for farmers in the development of the photovoltaic industry and other new energy industries.

Materials and methods

Theoretical basis

Double-bound dichotomous CVM

According to the stochastic utility maximization principle, the data on double-bound dichotomous CVM are analyzed by referring to the stochastic utility function model established by Hanemann (1989). The stochastic utility function model estimates the utility function through the bid value and its answer response data.

Assuming that the state of obtaining and maintaining the improved value of agricultural production resources and the environment changes from $Q_0$ to $Q_1$, the measured equivalent surplus value is used to evaluate the cost paid by farmers to obtain the value. If the answer to a respondent’s WTP is “Yes,” it means that he is willing to pay a given price for obtaining and maintaining the sustainable existence (existence value), future generations’ use (bequest value), and his own use (option value) of the photovoltaic industry for the improvement of agricultural production resources and the environment.

The income of the interviewee is $M$, and the indirect utility function of the interviewee is composed of two parts: the observable part $V$ and unobservable part $\epsilon$. Then, the utility function is $U=V\left(Q,M\right)+\epsilon$. The observable utility difference is denoted by $\Delta V$. When the bid value given to the respondent is $T$ yuan, the probability of the respondent answering “Yes” is $P\left[Y\mathrm{e}\mathrm{s}\right]={\left[1+e^{-∆V}\right]}^{-1}$. For the observable utility difference $\Delta V$, respondents may be affected by their socioeconomic characteristics and bid value. The linear function of $\Delta V$ is expressed as follows: $\Delta V=\alpha +\beta x_i+\gamma T$.

The core question of the double-bound dichotomous CVM first provides an initial bid value of $R_i$ to the respondents and asks them whether they are willing to pay. If the respondents answer “Yes,” another higher bid value of $R_i^u$ is provided to them. Otherwise, they are offered another lower bid value of $R_i^d$. Therefore, there are four possibilities of the answers of respondents: “Yes/Yes,” “Yes/No,” “No/Yes,” and “No/No.” $d_i^{YY}$, $d_i^{YN}$, $d_i^{NY}$, and $d_i^{NN}$ are dummy variables corresponding to the answer results of respondents “Yes/Yes,” “Yes/No,” “No/Yes,” and “No/No,” respectively. If the respondent’s answer is “Yes/Yes,” then $d_i^{YY}=1$; then, other $d_i^{YN}$, $d_i^{NY},$ and $d_i^{NN}$ are all 0.

Parameters $\alpha$, $\beta$, and $\gamma$ are estimated by maximum likelihood estimation. The expression of its likelihood estimation function is as follows:

$\ln \mathrm{L}={\displaystyle \sum }_{i=1}^{\mathrm{n}}\left[d_i^{YY}\ln P_{YY}+d_i^{YN}\ln P_{YN}+d_i^{NY}\ln P_{NY}+d_i^{NN}\ln P_{NN}\right].(1)$

The mean value of the WTP is calculated as follows:

$WT{P}_{mean}={\int }_0^{T_{\max }}\frac{dt}{1+\exp \left(-\alpha -\beta \overline{X}-cT\right)}(2)$

where $\alpha$ is a constant, $\gamma$ is the regression coefficient of the bid value, X^— is the average value of socioeconomic characteristic variables, and $\beta$ is the regression coefficient of socioeconomic characteristic variables.

Embedding effects

According to the consumption theory, the individual preference utility function is constructed. Suppose the initial supply level of the public goods is $S_0$, $y$ is the disposable income and $U\left(S_0,y\right)=U\left(S_0+a,y-WT{P}_a\right)$. It is assumed that the whole $a$ of the public goods to be estimated consists of a first increase part b and a second increase part c, so a+b = c. At independent valuation, the WTP of $a$, $b$ , and $c$ is $WT{P}_i^n$ ($i$ is $a$, $b$, and $c$, and $n$ is the number of independent valuations). Then, the utility function of the first increased part of the public goods to be estimated is $U\left(S_0,y\right)=U\left(S_0+b,y-WT{P}_b\right)$. In the case of $b$ already being possessed, the WTP for $c$ is $WT{P}_c|b$. Due to possible substitution and/or income effects (Whitehead, 2016), the utility function of $c$, the second additional part of the public goods to be evaluated is $U\left(S_0+b,y-WT{P}_b\right)=U\left(S_0+b+c,y-WT{P}_b-WT{P}_c|b\right)$. In field experiments, the income effect usually cannot be adjusted, so there is $U\left(S_0+b,y\right)=U\left(S_0+b+c,y-WT{P}_c|b\right)$.

If there are four cases of ① $WT{P}_a\le WT{P}_b+WT{P}_c|b$; ② $WT{P}_c|b=WT{P}_a-WT{P}_b\le WT{P}_c$; ③ $WT{P}_a\approx WT{P}_b$; and ④ $WT{P}_i^1\ne WT{P}_i^2\ne \cdots \ne WT{P}_i^n$ in the total test and partial test, it indicates the existence of an embedding effect. Considering the possibility and accuracy of the experiment, the scope test is usually carried out mainly to test whether there are two cases (③ and ④) (Kahneman and Knetsch, 1992; Boyle et al., 1994; Whitehead, 2016; Baron, 2017).

NOAA recommends the scope test as the standard method for assessing the internal consistency of CVM results. The scope test is a construct validity test based on the basic principle of the “more goods are better than less” consumption theory (Arrow, 1993; Borzykowski et al., 2018), which is usually divided into the internal scope test and external scope test (Carson and Mitchell, 1995). The internal scope test is to measure the WTP change of the same respondent under different ranges of public goods to be estimated, corresponding to the paired sample. The external scope test measures the WTP changes of different respondents in different ranges of public goods to be estimated, corresponding to sub-samples (Carson et al., 2001; Bateman et al., 2004). The internal scope test and the external scope test are complementary (Ndambiri et al., 2017).

Methods

Sample selection

In terms of green energy development, the Ningxia Hui Autonomous Region has high-quality natural conditions and is the country’s first comprehensive demonstration area for new energy. As of August 2022, the installed PV capacity of the province was 14.36 million kW. New energy generation accounted for 23% of the province’s total power generation, ranking second in the country. Ningxia is also the first province in China to combine solar photovoltaic technology with a facility agricultural greenhouse. As one of the earliest counties in Ningxia to develop a photovoltaic industry, Pingluo County has made outstanding contributions to the development of the clean energy industry and the improvement of green resources and the environment in the agricultural industry. Until now, the total installed capacity of clean energy industries in Pingluo County accounts for 49.6% of the total installed capacity of photovoltaic and wind power in the city. In conclusion, this study selects Pingluo County in Ningxia Province for agricultural technology development, construction, and management of the business of the photovoltaic industry. Four research sites are selected within the county for sampling, such as the photovoltaic industry area (Gaozhuang), two adjacent photovoltaic industry areas (Huang Quqiao and Touzha), and away from the photovoltaic industry area (Chonggang) (Figure 1), as a case study of farmers’ WTP for the photovoltaic industry to improve agricultural green resources and the environment.

FIGURE 1

FIGURE 1. Study area map.

Research design

Bid value design of the double-bound dichotomous CVM

The double-bound dichotomous CVM can simulate market transactions realistically, accurately describe the false market, reduce the bias caused by market distortions, and find a more realistic WTP for respondents (Changlin et al., 2016). In the given virtual situation, this study directly asked respondents about the WTP (Ressurreio et al., 2012; Fitzpatrick et al., 2017) of maintaining the existing photovoltaic industry to improve the state of agricultural green resources and the environment without decreasing (Ramdas and Mohamed, 2014) so as to ensure the sustainable existence (existence value) of these resources and the environment, the use of future generations (bequest value), and their future use (option value). As a means of eliciting WTP, the double-bound dichotomous CVM for each volunteer designs three bid values: a higher bid value ($R_i^u$), a lower bid value ($R_i^d$), and the initial bid value ($R_i$); according to the respondents’ bid value, there are four answers: $\left(Yes,Yes\right)$, $\left(Yes,No\right)$, $\left(No,Yes\right)$, and $\left(No,No\right)$. Based on the functional relationship between the probability of acceptance by the respondents and the bid price, the WTP of the respondents is derived (Whan et al., 2013).

In order to avoid the problem of bias in the bid value range in the double-boundary dichotomous CVM questionnaire, this study selects the C-optimal design based on Anna (1995), Pukelsheim (2007), and Kim (2009) to correct the bid value design, correctly estimate the bid value range, and reduce the bias due to inclusion of low bid values.

Suppose there is one and only one bid value scheme, then the initial value of the bid is equal to the median of the WTP derived from the pre-survey. The higher bid value and the lower bid value are distributed about the axis of symmetry of the initial bid value. Asymptotic variance minimization of the WTP median is the essence of the C-optimal design (Kim, 2009).

Based on the model construction of the double-bound dichotomous CVM, using maximum likelihood estimation, set $\theta =\left(\beta ,\sigma \right)$. $\exp \left(\mu \right)$ is set as the estimated median value of the WTP in this study, and $\exp \left(\mu \right)$ and $\mu$ have the same monotonicity. Then, the asymptotic variance expression of $\mu$ is

$Asy.Var\left(\widehat{\mu }\right)=\frac{I_{\mu \mu }}{\mathit{Det}\left[I\left(\mu ,\sigma \right)\right]}=\frac{4{\sigma }^2{\left(1+e^w\right)}^2}{N\left(-1+4e^w+e^{2w}\right)}.(3)$

N is the sample size, and the bid value is calculated according to the principle of asymptotic variance minimization and the expression: $w=(\log \left(R_i^d\right)-\mu )/\sigma =-(\log \left(R_i^{\mu }\right)-\mu )/\sigma$.

Embedded effect verification scheme design

Referring to the research results on the value evaluation of embedded public goods and serial public goods, this study adopts the exclusive list method (Bateman et al., 2001; Bateman et al., 2004); information disclosure adopts advanced disclosure design (Powe and Bateman, 2003); and the sequence of valuation questions adopts a combination of top–down and bottom–up methods (Carson and Mitchell, 1995; Halvorsen, 1996; Hanley et al., 2003; Mcdaniels et al., 2003). Research on the embedded effects of farmers’ WTP for resource and environmental improvement has been carried out.

From the perspective of the quantitative scope embedding effect, five groups of independent sub-samples were designed, and each group of independent sub-samples was designed with 1–3 core valuation questions about farmers’ WTP for the photovoltaic industry to improve agricultural resources and the environment. The region where the photovoltaic industry is located (Gaozhuang) is the core component of farmers’ WTP. Therefore, this study selected Gaozhuang, Huang Quqiao, Touzha, and Chonggang as the four research points, respectively, as the variation of the quantitative scope (Table 1).

TABLE 1

TABLE 1. Design ideas of the verification scheme for the quantitative scope embedding effects.

The upper corner marks “1–3″ represent the serial number of the payment value within the sub-sample. $AS$ represents the WTP for the improvement of resources and the environment, and $AS=gr+hr+tr+cr$ and $gr、hr、tr、cr$ represent the resources and environment to be improved in Gaozhuang, Huang Quqiao, Touzha, and Chonggang, respectively. $gr+hr$, $gr+hr$, and $gr+cr$ represent the resources and environment to be estimated in the pairwise combination of the region where the photovoltaic industry is located (Gaozhuang) and the rest of the survey area.

According to the design ideas of the verification scheme of quantitative scope embedding effects in Table 1, this study proposes the null hypothesis of the external quantitative scope embedding effect test and the internal quantitative scope embedding effect test of farmers’ WTP for the photovoltaic industry to improve agricultural production resources and the environment.

The null hypothesis of the external quantitative scope embedding effect test is as follows:

①  $ex-H_1:WTP\left[A{S}_{\mathrm{I}}^1\right]=WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right]=WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right]=WTP\left[A{S}_{\mathrm{I}\mathrm{V}}^1\right]=WTP\left[A{S}_{\mathrm{V}}^1\right],$

② $ex-H_2:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2],WTP[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right]\le WTP\left[A{S}_{\mathrm{I}}^1\right],$

③ $ex-H_3:WTP\left[h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2],WTP[(gr+tr{)}_{\mathrm{I}\mathrm{V}}^3\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right],$

④ $ex-H_4:WTP\left[t{r}_{\mathrm{I}\mathrm{V}}^2],WTP[(gr+cr{)}_{\mathrm{V}}^3\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right],$

⑤ $ex-H_5:WTP\left[c{r}_{\mathrm{V}}^2\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{V}}^1\right],$

⑥ $ex-H_6:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\le WTP\left[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right],WTP\left[(gr+tr{)}_{\mathrm{I}\mathrm{V}}^3\right],WTP\left[(gr+cr{)}_{\mathrm{V}}^3\right].$

The null hypothesis of the internal quantitative scope embedding effect test is as follows:

① $in-H_1:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right],$

② $in-H_2:WTP\left[h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2\right]\le WTP\left[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right],$

③ $in-H_3:WTP\left[t{r}_{\mathrm{I}\mathrm{V}}^2\right]\le WTP\left[(gr+tr{)}_{\mathrm{I}\mathrm{V}}^3\right]\le WTP\left[A{S}_{\mathrm{I}\mathrm{V}}^1\right],$

④ $in-H_4:WTP\left[c{r}_{\mathrm{V}}^2\right]\le WTP\left[(gr+cr{)}_{\mathrm{V}}^3\right]\le WTP\left[A{S}_{\mathrm{V}}^1\right].$

Questionnaire design and survey implementation

In this study, the double-bound dichotomous CVM question format was used to design the questionnaire, and the 15 guiding principles of the non-use value assessment proposed by NOAA were referenced (Arrow, 1993). On the basis of fully understanding the development status of the photovoltaic industry in Pingluo County, the main contents and methods of the questionnaire were determined through expert consultation. In order to improve the quality of the survey, field training was conducted for the investigators, and a pre-survey was organized in the sampled areas. Finally, according to the pre-survey results, the survey scheme was improved and the bidding value of the questionnaire was revised to improve the content and quality of the questionnaire design. In order to ensure the truth and accuracy of the data, this survey adopts the form of a one-on-one interview. Residents of the village and non-residents of the village were selected as the main respondents affected by the sample sampling. Farmers living in the villages of each survey site (Gaozhuang, Huang Quqiao, Touzha, and Chonggang) were the main sources of the sample of residents. The sampling frequency of farmers interviewed at home was higher. The main sources of non-village residents were farmers in the core blocks of each survey site and those participating in the market.

① Deviation control: Control the hypothetical bias by explicitly reminding respondents of budget constraints, adding additional costs and alternatives, and using anonymity (Johnston, 2006) and strictly training investigators. Control information deviation through pre-information disclosure and pre-investigation (Spash, 2002; Bateman et al., 2004). Adopt a payment card-based guidance method to ease the cognitive burden of respondents on the improvement of agricultural resources and the environment (Ressurreio et al., 2012). N initial bid values will correspond to N payment schemes. This study sets N+2 bid values in the questionnaire design, combined with the C-optimal design, to determine the final bid value of the questionnaire, reduce the deviation, and truncate lower bid values.

Based on the previous research, the random utility model is calculated in the double-bound dichotomous CVM model, assuming that the WTP obeys the logistic distribution, and the bid value is calculated: $\left(R_i^d,R_i^u\right)=\left[\exp \left(\widehat{\mu }-1.09861\widehat{\sigma }\right),\exp \left(\widehat{\mu }+1.09861\widehat{\sigma }\right)\right]$, so $\left(\widehat{\mu },\widehat{\sigma }\right)=\left(\mathrm{3.364,1.836}\right)$. Combined with the C-optimal design method, the final calculation is $\left(R_i^d,R_i^u\right)=\left(3.849,217.437\right)$. Considering the research basis of the existing value evaluation and the design of the bid value in the existing CVM research and taking into account the actual research situation and the truncation of the bid value, the final bid value is determined as $R_i^d=5$ and $R_i^u=2009$ (Table 2).

② Questionnaire structure: This study classifies the socioeconomic characteristics of farmers. In addition to the personal characteristics of farmers’ gender, age, and degree of education, it also covers farmers’ family characteristics (management type, labor force, and acreage), family property (household incomes per capita, home value, value of cultivated land, and household savings), and social capital (family members or friends serve as village cadres), which enhances the accuracy of calculating the WTP. Furthermore, this study measures the influencing factors of subjective environmental behavior assessment of farmers’ WTP for the photovoltaic industry to improve the agricultural production environment from three aspects: cognitive degree (know the impact of the photovoltaic industry on the environment), pay incentive (the photovoltaic industry is beneficial to the improvement of agricultural green resources and the environment), and behavior attitude (to support the photovoltaic industry in some way to improve the agricultural production environment), as shown in Table 3.

TABLE 2

TABLE 2. Design scheme before and after bid value revision.

TABLE 3

TABLE 3. WTP distribution of the double-bound dichotomous CVM.

The survey questionnaire adopts the combination of the double-bound dichotomous CVM and an embedded scheme. There are seven payment schemes for the bid value in the questionnaire and nine bid values for WTP. A total of five independent sub-sample questionnaires were designed.

The key question of the double-bound dichotomous CVM is that, according to the revised payment standard in Table 2, respondents are asked whether they are willing to pay relative to the initial payment price $R_i$ if the price is set to increase to $R_i^u$ (or decrease to $R_i^d$).The key question of the embedded scheme takes “S-sub-sample III” in Table 1 as an example, and then, the core valuation question about $A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1$, $h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2$, and $(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3$ is “in order to support the development of the local (Pingluo County/Huang Quqiao County/Gao Zhuang and Huang Quqiao County) photovoltaic industry and make full use of photovoltaic resources into agricultural production and farmers’ lives so that it can be sustainable and used by future generations or their own future use; it is envisaged to establish a development and protection fund if the fund is in the fundraising stage. Would you like to support this agricultural production environmental protection activity through a one-time donation, and would you like to pay yuan?“ The questionnaire is shown in Supplementary Material.

Descriptive statistical analysis

Zero payment for protest.

During the survey period, 2056 valid questionnaires were interviewed, and the reasons for no WTP, such as being uninterested in WTP or resource protection, rejecting donations or fund forms, and having nothing to do with oneself, were identified as protest zero payment (Dziegielewska and Mendelsohn, 2007; Lo and Jim, 2015; Khan et al., 2019b) in the protest payment questionnaire. There are 324 questionnaires, and the positive payment rate of these questionnaires is 84.2%. Residents and non-residents of the village samples accounted for 70.67% and 29.33%, respectively.

The valid data collected by the survey were classified and sorted, and the responses to the core questions in the double-bound dichotomous CVM questionnaire were obtained, as shown in Table 4.

TABLE 4

TABLE 4. Respondents’ socio-demographic and willingness to pay statistics.

According to the respondents’ answers to the core questions of the CVM questionnaire, we found that the respondents’ responses to the survey area followed the increasing bid value, and the probability of answering “No” also gradually increased.

Relationship and influence between the socio-demographic statistics and willingness to pay

Furthermore, the original continuous variables (acreage, household incomes per capita, home value, and value of cultivated land) were designed according to their quintiles or three quantiles (labor force) to form ordinal variables, to analyze the individual social attribute variables of the respondents in the valid questionnaire, and to increase the relationship and influence between the socio-demographic statistics of the respondents and their environmental behavior of the WTP for the improvement of agricultural resources and the environment, as shown in Table 4.

It can be seen that the difference in gender, age, and education degree has little influence on the WTP. However, farmers with a lower age and a higher education degree have a stronger awareness of ecological and environmental protection and could better understand and accept the importance of the photovoltaic industry to improving agricultural green resources and the environment, and have a higher WTP. If farmers rely more on agricultural production, they will pay more attention to the improvement of agricultural resources and the environment, and they will be more willing to invest. When the acreage of farmers increases, they can feel the impact of optimization and upgrading of resources and the environment on agricultural production more personally. Therefore, farmers with more acreage are more willing to pay. Household savings and home values are both positively correlated with WTP. Social capital has a strong positive impact on the WTP. Although the overall proportion of farmers with social capital, as defined in this study, is not high, they have the highest proportion of the WTP. To some extent, social capital provides a convenient way for farmers to access agricultural production policies and technological updates. As for the subjective factors affecting the WTP, from the perspectives of cognition, behavior, and attitude, most people are not active in the photovoltaic industry’s improvement of the agricultural environment, showing an “indifferent attitude.” The possible reason is that farmers in western China are in a relatively backward area and have no understanding of clean energy. It has not realized the promotional benefits of new energy for the promotion of agricultural industry technology and environmental improvement. However, it can be seen that the more the farmers tend to “agree” with the photovoltaic industry on the improvement of the agricultural environment in behavioral attitude and emotional cognition, the higher their WTP is. This analysis provides the basis for the following empirical analysis.

The significance test of the socio-demographic characteristic difference

Multiple independent-sample parametric one-way ANOVA and multiple independent-sample non-parametric Kruskal–Wallis test (Christie, 2001; Norinder et al., 2001) were used to test the significance of the differences in the socio-demographic characteristics in five independent subsamples (Khan et al., 2022). The results are shown in Table 5. From the test results, a significance level of 0.05 indicates that there is no significant difference in socio-demographic characteristics among the five independent subsamples, indicating that the questionnaire design and survey process are reliable.

TABLE 5

TABLE 5. Significance tests of socio-demographic differences.

Empirical results and discussion

Willingness to pay estimation

Using the data from the double-bound dichotomous core question to sort out the dependent variable data, the respondent’s personal social attribute variable is set as the independent variable and Logit regression is used to obtain the relationship expression between WTP and influencing factors as follows:

$\log \left(WT{P}_{mean}\right)={\beta }_0+{\beta }_1{G}_i+{\beta }_2{A}_i+{\beta }_3{E}_i+{\beta }_{4}M{T}_i+{\beta }_{5}L{F}_i+{\beta }_{6}A{c}_i+{\beta }_{7}H{I}_i+{\beta }_{8}H{V}_i+{\beta }_{9}V{L}_i+{\beta }_{10}H{S}_i+{\beta }_{11}V{C}_i+{\beta }_{12}{R}_i.$

On the basis of obtaining the average value of each variable and the coefficient of each variable, the average WTP of respondents under the dichotomous guidance technique can be obtained. According to the sorted survey data, the forward: LR method is used to gradually regress the model, and after eliminating the insignificant variables, the regression analysis (Table 6) can be obtained.

TABLE 6

TABLE 6. Estimated results of WTP for respondents of the double-bound dichotomous CVM.

In this study, a Hosmer–Lemeshow goodness-of-fit test was conducted on the model, and the survey data were randomly divided into 10 groups. The Sig of the display test was 0.772 > 0.5, so the model could fit the data well.

According to the regression results, in the WTP model, social capital, household savings, labor force, acreage, management type, and home value have a positive and significant impact on respondents’ WTP. According to the results of this analysis, the WTP increased with the number of family members working as village officials and the amount of savings. Household savings have a positive and significant impact, while household per capita income has not passed the significance test. It can be seen that savings among the influencing factors can better reflect the wealth level of respondents in the survey area.

In comparing the literature on the WTP for solar industry development and environmental protection (Jin et al., 2019; Ali et al., 2020; Khan et al., 2020b; Haque et al., 2022), the wealth benefits (income, house-ownership appear, etc.) of the respondents have a significant positive effect on the WTP, which is consistent with this study. However, education, as the main influencing factor in some literature reports, has not passed the test in this study. This is due to the fact that the education level of rural households in western China is generally lower than that of other regions, especially that of urban residents, and their awareness and acceptance of resource and environmental protection and green energy utilization are lower (Khan et al., 2019a; Khan et al., 2020a). Compared with most literature reports that take urban residents as the research object, this study chooses the influencing factor which is different from urban research and has agricultural characteristics like acreage and management type. At the same time, social capital, which has a significant impact on farmers’ behavior and moral hazard, is included in the consideration scope to further reduce the regression error. The results show that these variables pass the regression test.

In addition, this study uses OLS regression and interval regression (Cameron and Huppert 2006; Stefania, 2018) to calculate the willingness of respondents to pay for the improvement of agricultural production resources and the environment in the photovoltaic industry again and analyze the embedded effects. Due to the limitation to sample size (a total of 2,056 valid questionnaires divided into five independent subsamples), this study introduced dummy variables into OLS and IR (Henrik 2000) to consider the differences in socio-demographic characteristics and payment preferences between residents non-residents of the village. The WTP in this study is based on the C-optimal design of the bid value and other deviation control. OLS regression is suitable for estimation of the dependent variable with established boundaries (Cheng et al., 2022). It also avoids excessive correction that may distort the regression results. In order to better retain more information and robust regression results and improve the comparison effect, this study chooses IR regression, which is more suitable for the double-bound dichotomous CVM, and takes the payment interval as the dependent variable for regression. Three kinds of WTP estimation results are combined to improve the robustness and effectiveness of subsequent embedding effect analysis.

Table 7 shows the original mean, OLS mean, and IR mean of the 12 kinds of farmers’ WTP for the improvement of agricultural production resources and the environment.

TABLE 7

TABLE 7. Estimation results of the mean WTP.

It can be seen that the five groups of independent sub-samples estimated the WTP ($A{S}_{\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{V}}^1$; $A{S}_{\mathrm{V}}^1$) of farmers to protect the agricultural production resources and environmental improvement of the photovoltaic industry in the same area, and the original mean, OLS mean, and IR mean of WTP were relatively stable. The maximum difference in WTP among the five independent sub-samples of the three valuation methods was all less than 4%. At the same time, the change direction of the original mean, OLS mean, and IR mean of each WTP is consistent with the change direction of the quantitative scope; the WTP mean of the largest range > the WTP mean of the middle range > the WTP mean of the smallest range.

This study draws on the practice of Cameron and Huppert (2006) and Tonin (2019) and adopts OLS and IR regression to estimate WTP. Due to the limitation to sample size (a total of 2056 valid questionnaires divided into five independent subsamples), this study introduced dummy variables into OLS and IR (Henrik 2000) to consider the differences in socio-demographic characteristics and payment preferences between residents non-residents of the village (related contents have been added to the survey implementation and descriptive statistics in the article). Therefore, OLS and IR regression are more suitable for the research content of this study. Second, the WTP is based on the C-optimal design of the bid value and other deviation control, which reduces the error caused by the inclusion of a low bid value and also reduces the degree of data dispersion in regression analysis (Kim, 2009). Compared with the Tobit model that requires the explained variable to be “Left censor” and/or “Right censor,” OLS regression is more suitable for estimation of the dependent variable with established boundaries in this study (Cheng et al., 2022). It also avoids excessive correction that may distort the regression results. OLS regression is chosen to highlight the comparative results of WTP estimation. Compared with other regression methods, OLS is primitive and simple (Cameron and Huppert 2006; Cheng et al., 2022). It selects the midpoint value of each interval as the true WTP of the respondents as the dependent variable. Therefore, in order to better retain more information and robust regression results and improve the comparison effect, this study chooses interval regression, which is more suitable for the double-bound dichotomous CVM, and takes the payment interval as the dependent variable for regression. Three kinds of WTP estimation results are combined to improve the robustness and effectiveness of subsequent embedding effect analysis.

Analysis of scope test results

Internal scope test

This study uses two paired sample parametric t-tests and two paired sample non-parametric t-tests. The Wilcoxon signed-rank test compares two sets of paired samples. Using multiple paired sample non-parametric test in the Friedman test pair, three groups of paired samples were compared.

External scope test

This study uses a two independent sample parametric t-test and a two independent sample non-parametric Mann–Whitney U test to compare two independent samples (Khan et al., 2022). Multiple independent-sample parametric one-way ANOVA and multiple independent-sample non-parametric Kruskal–Wallis test (Christie, 2001; Norinder et al., 2001) were used to compare three independent samples and four independent samples. Due to the possible existence of the phenomenon of the “over-embedding effect” (Powe and Bateman et al., 2004; Heberlein et al., 2005; Frontuto et al., 2017), this study uses a two-tailed test for comparison between pairs, and the test results are shown in Table 8 and Table 9.

TABLE 8

TABLE 8. Results of internal scope tests.

TABLE 9

TABLE 9. Results of external scope tests.

According to the results of the scope test of embedding effect verification, the comparison between farmers’ WTP for the improvement of agricultural resources and the environment in the Huang Quqiao area and farmers’ WTP in the photovoltaic industry location (Gaozhuang) and Huang Quqiao area ($WTP\left[h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2\right]=WTP\left[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right]$) fails to pass the internal scope test of two paired samples. WTP comparisons of null tests for other hypotheses passed both internal scope parametric and non-parametric tests (Table 7).

By comparing the WTP for the improvement of agricultural production resources and the environment ($A{S}_{\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{V}}^1$; $A{S}_{\mathrm{V}}^1$) of farmers in the same region surveyed by different independent sub-samples, the WTP results for the maximum range of the five independent sub-samples were stable, which passed the external scope test, proving the reliability of the questionnaire design, survey implementation, and estimated WTP results. The comparison of farmers’ WTP for the improvement of agricultural resources and the environment between the location of the photovoltaic industry (Gaozhuang) and Pingluo County as a whole ($WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]=WTP\left[A{S}_{\mathrm{I}}^1\right]$) has passed the non-parametric test of external scope. The WTP comparisons of null tests for other hypotheses did not pass the external scope parametric and non-parametric tests (Table 8). It can be seen that respondents can perceive the changes in the internal or external quantitative scope of the photovoltaic industry improving agricultural resources and the environment, and the direction of the change in the WTP size is consistent with the direction of the change in the quantitative scope size (see following equations).

① Comparison of WTP for internal quantitative scope changes

$in-H_1:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\stackrel{19}{\to }WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right]$

$in-H_2:WTP\left[h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2\right]\stackrel{6}{\to }WTP(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3]\stackrel{10}{\to }WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right]$

$in-H_3:WTP\left[t{r}_{\mathrm{I}\mathrm{V}}^2\right]\stackrel{19}{\to }WTP[(gr+tr{)}_{\mathrm{I}\mathrm{V}}^3]\stackrel{17}{{\displaystyle \to }}WTP\left[A{S}_{\mathrm{I}\mathrm{V}}^1\right]$

$in-H_4:WTP\left[c{r}_{\mathrm{V}}^2\right]\stackrel{6}{\to }WTP\left[(gr+cr{)}_{\mathrm{V}}^3\right]\stackrel{9}{{\displaystyle \to }}WTP\left[A{S}_{\mathrm{V}}^1\right]$

Note: The number on the arrow is the marginal WTP, which is the same as follows.

② Comparison of each WTP for external quantitative scope changes (0–1–4)

$ex-H_2:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\stackrel{18}{\to }WTP\left[A{S}_{\mathrm{I}}^1\right]$

$ex-H_3:WTP\left[h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2\right]\stackrel{32}{\to }WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right]$

$ex-H_4:WTP\left[t{r}_{\mathrm{I}\mathrm{V}}^2\right]\stackrel{18}{\to }WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right]$

$ex-H_5:WTP\left[c{r}_{\mathrm{V}}^2\right]\stackrel{32}{\to }WTP\left[A{S}_{\mathrm{I}\mathrm{V}}^1\right]$

Note: 0, 1, and 4 represent the number of sample areas, which is the same as follows.

③ Comparison of each WTP for external quantitative scope changes (0–2–4)

$ex-H_2:WTP\left[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right]\stackrel{17}{{\displaystyle \to }}WTP\left[A{S}_{\mathrm{I}}^1\right]$

$ex-H_3:WTP\left[(gr+tr{)}_{\mathrm{I}\mathrm{V}}^3\right]\stackrel{16}{{\displaystyle \to }}WTP\left[A{S}_{\mathrm{I}\mathrm{I}}^1\right]$

$ex-H_4:WTP\left[(gr+cr{)}_{\mathrm{V}}^3\right]\stackrel{9}{{\displaystyle \to }}WTP\left[A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1\right]$

④ Comparison of each WTP for external quantitative scope changes (0–1–2)

$ex-H_6:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\stackrel{1}{\to }WTP\left[(gr+hr{)}_{\mathrm{I}\mathrm{I}\mathrm{I}}^3\right[$

$ex-H_6:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\stackrel{0}{\to }WTP\left[gr+tr{)}_{\mathrm{I}\mathrm{V}}^3\right[$

$ex-H_6:WTP\left[g{r}_{\mathrm{I}\mathrm{I}}^2\right]\stackrel{3}{\to }WTP\left[(gr+cr{)}_{\mathrm{V}}^3\right[$

The primary economic question for the quantitative scope test is the degree of saturation (Bateman et al., 2004). Whether WTP increases depends on the saturation degree of the respondents to the supply level of the photovoltaic industry to improve agricultural production resources and the environment (Chilton and Hutchinson, 2000). From the comparison of the results of each WTP, the increased marginal WTP value of the internal quantitative scope changes was low (①). The marginal WTP increase in the quantitative scope changes is also low (②, ③, and ④). At the same time, the willingness of farmers to pay for the photovoltaic industry to improve agricultural production resources and the environment shows a marginal decline, and the utility of respondents tends to be saturated. Combined with previous studies, this study has shown that saturation degree (Hutchinson and Hutchinson 2003) and diminishing marginal utility (Veisten et al., 2004; Desvousges et al., 2012) are the main factors affecting the magnitude and range sensitivity of marginal WTP added value of agricultural green resources and environmental quantitative scope changes.

In Pingluo County, farmers’ WTP for the photovoltaic industry to improve agricultural production resources and the environment has convergence and substitution within the region and between the same external regions. The WTP of farmers in the location of the photovoltaic industry (Gaozhuang) and its adjacent sample areas (Huang Quqiao and Touzha) on the external quantity range of the photovoltaic industry to improve agricultural production resources and the environment is 1 and 0, respectively. Pairwise WTP comparison shows a substitution effect (④). Thus, combined with previous studies, the substitution effect (Nunes and Schokkaert, 2003; Desvousges et al., 2017) is the main factor affecting the magnitude of the marginal WTP increase and range sensitivity of quantitative scope changes. Farmers in the regions where the photovoltaic industry is located and radiation areas have a high awareness of the photovoltaic industry and an urgent desire to improve agricultural industry technology and production environments by using photovoltaic industry technology.

Conclusion, recommendations, limitations, and future directions

1) This study uses a combination of the double-bound dichotomous CVM and an embedding scheme to measure the farmers’ WTP for the photovoltaic industry to improve agricultural resources and the environment. After estimating the WTP for the improvement of agricultural resources and the environment of farmers in the four survey sites ($g{r}_{\mathrm{I}\mathrm{I}}^2$; $h{r}_{\mathrm{I}\mathrm{I}\mathrm{I}}^2$; $t{r}_{\mathrm{I}\mathrm{V}}^2$, and $c{r}_{\mathrm{V}}^2$), the sum of WTP is 304 yuan. For the overall valuation of farmers’ WTP in the four regions, the mean WTP of the five independent sub-samples ($A{S}_{\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{I}\mathrm{I}}^1$; $A{S}_{\mathrm{I}\mathrm{V}}^1$, and $A{S}_{\mathrm{V}}^1$) was 99.8 yuan, that is, in order to obtain and maintain the local photovoltaic industry to improve the agricultural production resources and environment, the average willingness of each respondent was to pay 99.80 yuan per year, respectively. The sum of the independent estimates of the four areas is 3.04 times the mean value of the overall estimates of the four areas. In addition, the resource endowment of farmers has a significant impact on their WTP, and the higher the social capital and financial capital of farmers, the more obvious the positive effect on the WTP.

2) In this study, five independent sub-sample questionnaires, including seven bidding value payment schemes, were designed, and four internal range tests and six external range tests were designed. The results of the range test show that the comparison of the WTP for the improvement of agricultural resources and the environment among different independent sub-samples in the same region has passed the external quantitative scope test, and the hypothesis $ex-H_1$ has passed, indicating that the questionnaire design, survey implementation, and WTP estimation results of the double-bound dichotomous CVM in this study are reliable. The comparison of WTP results with other external quantitative scope tests showed differences, indicating that the respondents were sensitive to changes in the external quantity scope of the photovoltaic industry for the improvement of agricultural production resources and the environment but did not reach statistical significance. Except for part of the internal scope test ($in-H_2$), all the other tests passed, indicating that the respondents were sensitive to the changes in the internal scope of the improvement of agricultural production resources and the environment to be estimated and reached statistical significance.

3) When the quantity range changes, the utility of respondents tends to saturate and marginal utility rapidly decreases. At the same time, there are substitutions between the WTP of farmers in some research areas, which accelerates the decline of the marginal WTP added value. The degree of saturation, diminishing marginal utility, or substitution effect are the main influencing factors of the magnitude change of the marginal WTP added value, which provides the basis for the external and internal quantitative scope tests to fail to reach statistical significance. Therefore, the range test should not be used as the only criterion to judge the validity of CVM research. The improvement of the photovoltaic industry to the agricultural green production environment and the effective use of photovoltaic technology in the optimization of agricultural green production technology are inseparable from the promotion of the government. When developing the photovoltaic industry, the local government should fully consider the substitution effect and convergence effect between different regions and comprehensively consider the saturation degree and diminishing marginal utility. Support policies for the photovoltaic industry should be formulated according to local conditions and be in line with the needs of agricultural green production so as to maximize the radiation benefits of green energy for the improvement of agricultural production resources and the environment.

4) Affected by factors such as saturation degree, diminishing marginal utility, and substitution effect, the overall valuation result of farmers’ WTP for the local photovoltaic industry to improve agricultural production resources and the environment is more reliable in cost-benefit analysis. According to the calculation results, it is suggested to choose 99.80 yuan in cost-benefit analysis. According to the farmers’ WTP, it provides an empirical reference for the development of photovoltaic industry compensation and subsidy measures for farmers to help increase farmers’ awareness of the photovoltaic industry’s improvement of agricultural green production resources and environment, increase farmers’ enthusiasm to support the development of the photovoltaic industry, and promote the input and utilization of photovoltaic technology in agricultural production. This conclusion also applies to the valuation of public goods in a complex environment.

Embeddings are broader economic phenomena, and failure to meet the range test should not be used as a preliminary basis for rejecting CVM studies, and the results of the range test should be considered more carefully. Scope insensitivity may be a problem with any assessment method, but current research has focused on the context of declarative preferences dominated by the CVM domain.

Further research may focus on the following discussions:

1) Research design is a key factor in determining the validity and reliability of the non-use valuation results of complex public goods. When designing the questionnaire, it is recommended to use the independent list method to evaluate the public goods to be estimated.

2) It may focus on the affective, cognitive, attitude, and behavior of social psychological theory factors on the embedding effect. The relationship between the warm glow and the embedding effect is controversial at present and is worth discussing in the future.

3) By breaking the limit of sample size, the embedded-effect problem of different resident characteristic samples was examined separately by increasing the sample size. Other types of embedding effects, such as the classification range embedding effect and geographic range embedding effect, are considered.

4) At present, the relationship between social psychological factors, warm glow, and the embeddedness effect is a hot and frontier direction in the research of CVM validity and reliability, which needs further exploration and study.

Data availability statement

The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.

Ethics statement

The studies involving human participants were reviewed and approved by Northwest A&F University. The patients/participants provided their written informed consent to participate in this study.

Author contributions

JL: conceptualization, formal analysis, investigation, software, methodology, data curation, writing—original draft, and writing—review and editing. AK: investigation, software, methodology, and writing—review and editing. SK: methodology and writing—review and editing. MA: data curation and writing—review and editing. JL: funding acquisition, project administration, and supervision. All authors contributed to the article and approved the submitted version.

Funding

This paper is supported by “Research on the Dynamic Value Evaluation of Agricultural Biological Assets, Mortgage Financing Model and Risk Management Policy”, National Natural Science Foundation of China (NSFC), Jan 2023-Dec 2026, No. 72273105. Sponsor and Host: JL. This paper is also supported by “Research on the Effectiveness Evaluation, Risk Control and System Construction of the Agricultural Credit Guarantee Policy”, National Natural Science Foundation of China (NSFC), Jan 2019-Dec 2022, No. 71873100. Sponsor and Host: JL. This paper is also supported by “Rural revitalization financial policy innovation team”, Chinese Universities Scientific Fund, Jan 2022-Dec 2023, No. 2452022074. Sponsor and Host: JL. This paper is also supported by “Research on the Policy Orientation and Implementation Path of Financial Empowerment of Rural Revitalization”, the Soft Science Project of the Central Agricultural Office and the Rural Revitalization Expert Advisory Committee of the Ministry of Agriculture and Rural Affairs, 2022.5.31-2023.5.31, No. rkx20221801, Sponsor and host: JL.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors, and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvs.2022.1029568/full#supplementary-material

References

Ali, M. A. S., Khan, S. U., Khan, A., Khan, A. A., and Zhao, M. (2020). Ranking of ecosystem services on the basis of willingness to pay: Monetary assessment of a subset of ecosystem services in the Heihe River basin. Sci. Total Environ. 734, 139447. doi:10.1016/j.scitotenv.2020.139447

PubMed Abstract | CrossRef Full Text | Google Scholar

Anna, A. (1995). Optimal designs for discrete choice contingent valuation surveys: Single-bound, double-bound, and bivariate models. J. Environ. Econ. Manag. 28, 287–306. doi:10.1006/jeem.1995.1019

CrossRef Full Text | Google Scholar

Arrow, K. (1993). Report of the NOAA panel on Contingent Valuation, Fed. Regist., 58, 4601–14.

Google Scholar

Baron, J. (2017). Contingent valuation: Flawed logic? Science 357, 363. doi:10.1126/science.aan6594

PubMed Abstract | CrossRef Full Text | Google Scholar

Bateman, I. J., Cole, M., Cooper, P., Georgiou, S., Hadley, D., and Poe, G. L. (2004). On visible choice sets and scope sensitivity. J. Environ. Econ. Manage. 47, 71–93. doi:10.1016/s0095-0696(03)00057-3

CrossRef Full Text | Google Scholar

Bateman, I. J., Cooper, P., Georgiou, S., and Poe, G. L. (2001). Visible choice sets and scope sensitivity: An experimental and field test of study design effects upon nested contingent values. Univ. east anglia.

Google Scholar

Borzykowski, , , Baranzini, A., and Maradan, D. (2018). Scope effects in contingent valuation: Does the assumed statistical distribution of WTP matter? Ecol. Econ. 144, 319–329. doi:10.1016/j.ecolecon.2017.09.005

CrossRef Full Text | Google Scholar

Boyle, K. J., Desvousges, W. H., Johnson, F. R., Dunford, R. W., and Hudson, S. P. (1994). An investigation of part-whole biases in contingent-valuation studies. J. Environ. Econ. Manage. 27, 64–83. doi:10.1006/jeem.1994.1026

CrossRef Full Text | Google Scholar

Brock, T., and Reinhilde, S. (2014). Biodiversity and key ecosystem services as protection goal for agro-ecosystems in Europe, Abstr. Pap. Am. Chem. Soc., 248.

Google Scholar

Cameron, H., and Huppert, D. D. (2006). OLS versus ML estimation of non-market resource values with payment card interval data. J. Environ. Econ. Manage. 17, 230–246. doi:10.1016/0095-0696(89)90018-1

CrossRef Full Text | Google Scholar

Carson, R., Flores, N., and Meade, N. (2001). Contingent valuation: Controversies and evidence. Environmental and Resource Economics 19, 173–210.

Google Scholar

Carson, R. T., Mitchell, R. C., Hanemann, W. M., Kopp, R. J., Presser, S., and Ruud, P. A. (1992). A contingent valuation study of lost passive use values resulting from the exxon valdez oil spill. Germany: University Library of Munich.

Google Scholar

Carson, R. T., and Mitchell, R. C. (1995). Sequencing and nesting in contingent valuation surveys. J. Environ. Econ. Manage. 28, 155–173. doi:10.1006/jeem.1995.1011

CrossRef Full Text | Google Scholar

Changlin, A., Ling, Z., Yang, J., and Shixue, W. (2016). Influence of initial bid number and sample size on the double-bounded dichotomous choice contingent valuation method. Acta eco. Sin. 36. 854-862. doi:10.5846/stxb201405191024

CrossRef Full Text | Google Scholar

Cheng, P., Dong, Y., Wang, Z., Tang, H., Jiang, P., and Liu, Y. (2022). What are the impacts of livelihood capital and distance effect on farmers' willingness to pay for coastal zone ecological protection? Empirical analysis from the beibu gulf of China. Ecol. Indic. 140, 109053. doi:10.1016/j.ecolind.2022.109053

CrossRef Full Text | Google Scholar

Chilton, S. M., and Hutchinson, W. G. (2000). A note on the warm glow of giving and scope sensitivity in contingent valuation studies. J. Econ. Psychol. 21, 343–349. doi:10.1016/s0167-4870(00)00009-x

CrossRef Full Text | Google Scholar

Christie, M. (2001). A comparison of alternative contingent valuation elicitation treatments for the evaluation of complex environmental policy. J. Environ. Manag. 62, 255–269. doi:10.1006/jema.2001.0439

PubMed Abstract | CrossRef Full Text | Google Scholar

Desvousges, W., Mathews, K., and Train, K. (2012). Adequate responsiveness to scope in contingent valuation. Ecol. Econ. 84, 121–128. doi:10.1016/j.ecolecon.2012.10.003

CrossRef Full Text | Google Scholar

Desvousges, W., Mathews, K., and Train, K. (2017). An adding-up test on contingent valuations of river and lake quality. Land Econ. 91, 556–571. doi:10.3368/le.91.3.556

CrossRef Full Text | Google Scholar

Dziegielewska, D. A., and Mendelsohn, R. (2007). Does "No" mean "No"? A protest methodology. Environ. Resour. Econ. (Dordr). 38, 71–87. doi:10.1007/s10640-006-9057-4

CrossRef Full Text | Google Scholar

Ferreira, S., and Marques, R. C. (2015). Contingent valuation method applied to waste management, Resour. Conservation Recycl., 99, 111–117. doi:10.1016/j.resconrec.2015.02.013

CrossRef Full Text | Google Scholar

Fitzpatrick, L., Parmeter, C. F., and Agars, J. (2017). Threshold effects in meta-analyses with application to benefit transfer for coral reef valuation. Ecol. Econ. 133, 74–85. doi:10.1016/j.ecolecon.2016.11.015

CrossRef Full Text | Google Scholar

Frontuto, V., Dalmazzone, S., Vallino, E., and Giaccaria, S. (2017). Earmarking conservation: Further inquiry on scope effects in stated preference methods applied to nature-based tourism. Tour. Manag. 60, 130–139. doi:10.1016/j.tourman.2016.11.017

CrossRef Full Text | Google Scholar

Giguere, C., Moore, C., and Whitehead, J. C. (2020). Valuing hemlock woolly adelgid control in public forests: Scope effects with attribute nonattendance. Land Econ. 96, 25–42. doi:10.3368/le.96.1.25

PubMed Abstract | CrossRef Full Text | Google Scholar

Grammatikopoulou, I., and Olsen, S. B. (2013). Accounting protesting and warm glow bidding in Contingent Valuation surveys considering the management of environmental goods – an empirical case study assessing the value of protecting a Natura 2000 wetland area in Greece. J. Environ. Manag. 130, 232–241. doi:10.1016/j.jenvman.2013.08.054

PubMed Abstract | CrossRef Full Text | Google Scholar

Halvorsen, B. (1996). Ordering effects in contingent valuation surveys. Environ. Resour. Econ. (Dordr). 8, 485–499. doi:10.1007/bf00357416

CrossRef Full Text | Google Scholar

Hanemann, W. M. 1989. Welfare evaluations in contingent valuation experiments with discrete response data: Reply. Am. J. Agric. Econ., 71, 1057–1061. doi:10.2307/1242685

CrossRef Full Text | Google Scholar

Hanley, N., and Czajkowski, M. (2019). The role of stated preference valuation methods in understanding choices and informing policy. Rev. Environ. Econ. Policy 13, 248–266. doi:10.1093/reep/rez005

CrossRef Full Text | Google Scholar

Hanley, N., Schlaepfer, F., and Spurgeon, J. (2003). Aggregating the benefits of environmental improvements: Distance-decay functions for use and non-use values. J. Environ. Manag. 68, 297–304. doi:10.1016/s0301-4797(03)00084-7

CrossRef Full Text | Google Scholar

Haque, M. N., Saroar, M., Fattah, M. A., and Morshed, S. R. (2022). Environmental benefits of blue ecosystem services and residents’ willingness to pay in Khulna city, Bangladesh. Heliyon 8, e09535. doi:10.1016/j.heliyon.2022.e09535

PubMed Abstract | CrossRef Full Text | Google Scholar

Hausman, D., and Hausman, J. A. (1994). Contingent valuation: Is some number better than No number? J. Econ. Perspect. 8, 45–64. doi:10.1257/jep.8.4.45

PubMed Abstract | CrossRef Full Text | Google Scholar

Hausman, J. (2012). Contingent valuation: From dubious to hopeless. J. Econ. Perspect. 26, 43–56. doi:10.1257/jep.26.4.43

CrossRef Full Text | Google Scholar

Heberlein, T. A., Wilson, M. A., Bishop, R. C., and Schaeffer, N. C. (2005). Rethinking the scope test as a criterion for validity in contingent valuation. J. Environ. Econ. Manag. 50, 1–22. doi:10.1016/j.jeem.2004.09.005

CrossRef Full Text | Google Scholar

Henrik, S. (2000). Contingent valuation of global environmental resources: Test of perfect and regular embedding. J. Econ. Psychol. 21, 605–623. doi:10.1016/s0167-4870(00)00022-2

CrossRef Full Text | Google Scholar

Hutchinson, S., and Hutchinson, W. (2003). A qualitative examination of how respondents in a contingent valuation study rationalise their WTP responses to an increase in the quantity of the environmental good. J. Econ. Psychol. 24, 65–75. doi:10.1016/s0167-4870(02)00155-1

CrossRef Full Text | Google Scholar

Jin, J., Wan, X., Lin, Y., Kuang, F., and Ning, J. (2019). Public willingness to pay for the research and development of solar energy in Beijing, China. Energy Policy 134, 110962. doi:10.1016/j.enpol.2019.110962

CrossRef Full Text | Google Scholar

Johnston, R. J. (2006). Is hypothetical bias universal? Validating contingent valuation responses using a binding public referendum. J. Environ. Econ. Manag. 52, 469–481. doi:10.1016/j.jeem.2005.12.003

CrossRef Full Text | Google Scholar

Jorgensen, S. L., Olsen, S. B., Ladenburg, J., Martinsen, L., Svenningsen, S. R., and Hasler, B. (2013). Spatially induced disparities in users' and non-users' WTP for water quality improvements—testing the effect of multiple substitutes and distance decay. Ecol. Econ. 92, 58–66. doi:10.1016/j.ecolecon.2012.07.015

CrossRef Full Text | Google Scholar

Kahneman, D., and Knetsch, J. L. (1992). Valuing public goods: The purchase of moral satisfaction. J. Environ. Econ. Manage. 22, 57–70. doi:10.1016/0095-0696(92)90019-s

CrossRef Full Text | Google Scholar

Kerry, V., Smith, A. N. D., Laura, L., and Osborne, (1996). Do contingent valuation estimates pass a "scope" test? A meta-analysis. J. Environ. Econ. Manage. 31, 287–301. doi:10.1006/jeem.1996.0045

CrossRef Full Text | Google Scholar

Khan, S. U., Guo, X., Hu, J., Khan, A. A., Talpur, M. A., Liu, G., et al. (2022). Who cares and how much? Narrative for advances in aquatic ecosystem services through non-market valuation with spatial dimensions using a discrete choice experiment. J. Clean. Prod. 337, 130603. doi:10.1016/j.jclepro.2022.130603

CrossRef Full Text | Google Scholar

Khan, S. U., Hayat, S., Xia, X., Liu, G., and Zhao, M. (2020a). Improvisation of indigenous environmental benefit transfer and valuation for cleaner environment: Choice experiment across northwest China. J. Clean. Prod. 274, 123176. doi:10.1016/j.jclepro.2020.123176

CrossRef Full Text | Google Scholar

Khan, S. U., Khan, I., Zhao, M., Chien, H., Lu, Q., Ali, M. A. S., et al. (2019a). Spatial heterogeneity of ecosystem services: A distance decay approach to quantify willingness to pay for improvements in heihe river basin ecosystems. Environ. Sci. Pollut. Res. 26, 25247–25261. doi:10.1007/s11356-019-05691-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Khan, S. U., Khan, I., Zhao, M., Khan, A. A., and Ali, M. A. S. (2019b). Valuation of ecosystem services using choice experiment with preference heterogeneity: A benefit transfer analysis across inland river basin. Sci. Total Environ. 679, 126–135. doi:10.1016/j.scitotenv.2019.05.049

PubMed Abstract | CrossRef Full Text | Google Scholar

Khan, S. U., Liu, G., Zhao, M., Chien, H., Lu, Q., Khan, A. A., et al. (2020b). Spatial prioritization of willingness to pay for ecosystem services. A novel notion of distance from origin’s impression. Environ. Sci. Pollut. Res. 27, 3100–3112. doi:10.1007/s11356-019-06538-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Kim, J. (2009). Optimal experimental design for double-bounded dichotomous choice contingent valuation. Land Econ. 69, 138–146. doi:10.2307/3146514

CrossRef Full Text | Google Scholar

Li, X., Shuo, G., Jiuchang, W., and Qiying, D. (2022). Corrigendum to “Government subsidy and corporate green innovation - does board governance play a role? Energy Pol. 168, 112720.

Google Scholar

Lo, A. Y., and Jim, C. Y. (2015). Protest response and willingness to pay for culturally significant urban trees: Implications for Contingent Valuation Method. Ecol. Econ. 114, 58–66. doi:10.1016/j.ecolecon.2015.03.012

CrossRef Full Text | Google Scholar

Lu, Y., Norse, D., and Powlson, D. (2020). Agriculture green development in China and the UK: Common objectives and converging policy pathways. Front. Agric. Sci. Eng. 7, 98. doi:10.15302/j-fase-2019298

CrossRef Full Text | Google Scholar

Mcdaniels, T. L., Gregory, R., Arvai, J., and Chuenpagdee, R. (2003). Decision structuring to alleviate embedding in environmental valuation. Ecol. Econ. 46, 33–46. doi:10.1016/s0921-8009(03)00103-4

CrossRef Full Text | Google Scholar

Ndambiri, H., Mungatana, E., and Brouwer, R. (2017). Scope effects of respondent uncertainty in contingent valuation: Evidence from motorized emission reductions in the city of nairobi, Kenya. J. Environ. Plan. Manag. 60, 22–46. doi:10.1080/09640568.2016.1140024

CrossRef Full Text | Google Scholar

Nizam, M. S., Rozana, Z., Wahid, O., Zaimi, A. M. M., Latif, S. A., Mushairry, M., et al. (2013). Potential of solar farm development at UTM campus for generating green energy. Appl. Mech. Mater., 479–480, 553-558. doi:10.4028/www.scientific.net/AMM.479-480.553

PubMed Abstract | CrossRef Full Text | Google Scholar

Norinder, A., Hjalte, K., and Persson, U. (2001). Scope and scale insensitivities in a contingent valuation study of risk reductions. Health Policy 57, 141–153. doi:10.1016/s0168-8510(01)00117-8

PubMed Abstract | CrossRef Full Text | Google Scholar

Nunes, P., and Schokkaert, E. (2003). Identifying the warm glow effect in contingent valuation. J. Environ. Econ. Manage. 45, 231–245. doi:10.1016/s0095-0696(02)00051-7

CrossRef Full Text | Google Scholar

Ojea, E., and Loureiro, M. L. (2010). Valuing the recovery of overexploited fish stocks in the context of existence and option values. Mar. Policy 34, 514–521. doi:10.1016/j.marpol.2009.10.007

CrossRef Full Text | Google Scholar

Pinto, R., Brouwer, R., Patricio, J., Abreu, P., Marta-Pedroso, C., Baeta, A., et al. (2016). Valuing the non-market benefits of estuarine ecosystem services in a river basin context: Testing sensitivity to scope and scale. Estuar. Coast. Shelf Sci. 169, 95–105. doi:10.1016/j.ecss.2015.11.028

CrossRef Full Text | Google Scholar

Powe, N. A., and Bateman, I. A. N. J. (2004). Investigating insensitivity to scope: A split-sample test of perceived scheme realism. Land Econ. 80, 258–271. doi:10.2307/3654742

CrossRef Full Text | Google Scholar

Powe, N. A., and Bateman, I. J. (2003). Ordering effects in nested 'top-down' and 'bottom-up' contingent valuation designs. Ecol. Econ. 45, 255–270. doi:10.1016/s0921-8009(03)00084-3

CrossRef Full Text | Google Scholar

Pukelsheim, F. (2007). On c-optimal design measures. Ser. Stat. 12, 13–20. doi:10.1080/02331888108801566

CrossRef Full Text | Google Scholar

Rahim, A. A. (2022). A new green revolution (GR) or neoliberal entrenchment in agri-food systems? Exploring narratives around digital agriculture (DA), food systems, and development in sub-sahara africa. J. Dev. Stud. 58, 1588–1604. doi:10.1080/00220388.2022.2032673

CrossRef Full Text | Google Scholar

Ramdas, M., and Mohamed, B. (2014). Impacts of tourism on environmental attributes, environmental literacy and willingness to pay: A conceptual and theoretical review. Procedia - Soc. Behav. Sci. 144, 378–391. doi:10.1016/j.sbspro.2014.07.307

CrossRef Full Text | Google Scholar

Ressurreio, A., Gibbons, J., Kaiser, M., Dentinho, T. P., Zarzycki, T., Bentley, C., et al. (2012). Different cultures, different values: The role of cultural variation in public's WTP for marine species conservation. Biol. Conserv. 145, 148–159. doi:10.1016/j.biocon.2011.10.026

CrossRef Full Text | Google Scholar

Riti, C., Kumar, A. S., Amitava, B., Anwesha, M., Tufleuddin, B., Srikanta, D., et al. (2021). Conservation agriculture in new alluvial agro-ecology: Differential perception and adoption. J. Rural Stud. 88, 14–27. doi:10.1016/j.jrurstud.2021.10.001

CrossRef Full Text | Google Scholar

Ronald, C., and Glenn, H. (1995). The measurement and decomposition of nonuse values: A critical review. Environ. Resour. Econ. (Dordr). 5, 225–247. doi:10.1007/bf00691518

CrossRef Full Text | Google Scholar

Schulze, W. D., Mcclelland, G. H., Lazo, J. K., and Rowe, R. D. (1998). Embedding and calibration in measuring non-use values. Resour. Energy Econ. 20, 163–178. doi:10.1016/s0928-7655(97)00034-1

CrossRef Full Text | Google Scholar

Spash, C. L. (2002). Informing and forming preferences in environmental valuation: Coral reef biodiversity. J. Econ. Psychol. 23, 665–687. doi:10.1016/s0167-4870(02)00123-x

CrossRef Full Text | Google Scholar

Stefania, T. A. (2018). Citizens' perspectives on marine protected areas as a governance strategy to effectively preserve marine ecosystem services and biodiversity. Ecosyst. Serv. 34, 189–200. doi:10.1016/j.ecoser.2018.03.023

CrossRef Full Text | Google Scholar

Tonin, S. (2019). Estimating the benefits of restoration and preservation scenarios of marine biodiversity: An application of the contingent valuation method. Environ. Sci. Policy 100, 172–182. doi:10.1016/j.envsci.2019.07.004

CrossRef Full Text | Google Scholar

Veisten, K., Fredrik Hoen, H., Navrud, S., and Strand, J. (2004). Scope insensitivity in contingent valuation of complex environmental amenities. J. Environ. Manag. 73, 317–331. doi:10.1016/j.jenvman.2004.07.008

PubMed Abstract | CrossRef Full Text | Google Scholar

Waldo, F. K., Saaymanandrea, M., and Saayman, A. (2016). Willingness to pay: Who are the cheap talkers? Ann. Tour. Res. 56, 96–111. doi:10.1016/j.annals.2015.11.009

CrossRef Full Text | Google Scholar

Whan, H. T., Yiseok, H., and Sug, P. C. (2013). Valuation of han river waterside landscape with a double-bound dichotomous choice model and policy implications: Focused on the exponential willingness to pay model. Environ. Resour. Econ. Rev. 22, 179–214. doi:10.15266/kerea.2013.22.1.179

CrossRef Full Text | Google Scholar

Whitehead, J. C. (2016). Plausible responsiveness to scope in contingent valuation. Ecol. Econ. 128, 17–22. doi:10.1016/j.ecolecon.2016.03.011

CrossRef Full Text | Google Scholar

Yu, X. (2020). Promoting Agriculture Green Development to realize the great rejuvenation of the Chinese nation. Front. Agric. Sci. Eng. 7, 112. doi:10.15302/j-fase-2019318

CrossRef Full Text | Google Scholar