Factors Influencing Resident Satisfaction with Afforestation in the Plains: Beijing as a Case Study

Abstract

The acceleration of global urbanization has brought the issue of environmental degradation to the forefront. To effectively curtail these issues, cities worldwide are promoting afforestation; however, only a few studies have investigated levels of satisfaction from the perspective of residents, who are the main beneficiaries of these afforestation projects. In this study, we used the Ologit model in conjunction with 1158 survey samples from Beijing to empirically analyze the level of resident satisfaction with the Plain Afforestation Project (PAP). Results showed that (i) landscaping after afforestation, (ii) ecological aspects, and (iii) availability of recreational space significantly improved the overall resident satisfaction with the PAP. Additionally, older people are more likely to be satisfied with the PAP compared to younger residents, and male residents are more likely to be satisfied with the PAP than female residents. Residents with higher education and income levels were less satisfied with the PAP. These results provide a reference for decision-makers to optimize the PAP.

1. Introduction

As urbanization continues to accelerate, urban ecological issues, such as air pollution, noise pollution, and biodiversity degradation, are becoming increasingly prominent [1,2,3]. Beijing, the political, economic, social, and cultural center of China, is facing various adverse environmental issues, such as sandstorms, dust, and smog, as well as the degradation of urban and suburban living environments [4,5]. In response, a major project targeting the afforestation of one million mu (1 hectare = 15 mu) in plain areas was proposed by the Beijing Municipal Party Committee and Municipal Government in 2012 [6,7]. The aims of the Plain Afforestation Project (PAP) are to improve the ecological environment around the city and the quality of the urban living environment via afforestation [8,9].

Urban greening is considered key to building a livable city. Urban afforestation projects involve ecologically important functions, such as air purification, biodiversity protection, and noise reduction, and aim to objectively improve the urban environment and quality of life of urbanites [10,11,12]. As micro-subjects of urban development, residents are the main beneficiaries of afforestation of plains; thus, the efficacy of project implementation is closely associated with the environmental perception and satisfaction of residents [13,14,15]. Deng et al. examined the similarities and differences in the perceptions on leisure in Washington’s urban forests among tourists, repeat visitors, and local residents and found that locations with diverse and dense urban forests were more likely to inspire positive perceptions [16]. Similarly, Tesler et al. found that urban forests significantly increased the impact of life satisfaction in Israel [17]. Ostoic et al. analyzed the perception of and satisfaction with urban forests and green spaces in seven cities in southeastern Europe and found that citizens are genuinely concerned about forests and green spaces in cities but are not satisfied with the status quo [18]. Therefore, surveying the direct environmental perception of residents and their satisfaction with the PAP is conducive to accurately evaluating the effectiveness of the PAP, which in turn will help clarify the shortcomings of the afforestation project, leading to its optimization.

Scholars have conducted extensive research on the ecological and economic benefits of urban afforestation projects [16,17,18,19]. For example, Xu et al. [4] analyzed the impact of forest city construction on urban air quality in the Beijing–Tianjin–Hebei region and found that forest city construction significantly reduces haze-linked pollution in the Beijing–Tianjin–Hebei region. He et al. [1] evaluated the impact of forest city construction on business performance. They found that forest cities improve the business performance of enterprises via labor productivity, capital allocation, innovation ability, and management efficiency. However, few studies have conducted a comprehensive evaluation of PAPs from the perspective of the personal perceptions of residents as an entry point [20,21]. Thus, project performance with respect to improving livelihood and other vital interests of residents remains unknown. Therefore, this study evaluated the personal perception of residents as the breakthrough point, utilizing survey data of resident satisfaction with the Beijing PAP as a basis, and empirically analyzed factors that affect resident satisfaction with the Beijing PAP.

2. Case Study of Beijing

Beijing is located in the transition zone extending from the North China Plains to the Loess Plateau and Inner Mongolia Plateau in the northwest. The total area of the plains accounts for approximately 38% of the total area of Beijing, and most of the city’s population is located here. The plain area, which faces the Beijing–Tianjin–Hebei region and radiates to the Bohai Rim region and is intended for the urban expansion of the capital region, functions as an important area for bearing Beijing’s increasing economic aggregation, helps reduce the population of the central city, and provides an external window and a developmental frontier for Beijing [22,23].

To improve the ecological environment of the capital, ensure its ecological security, promote the construction of an ecologically appropriate civilization in the capital, and build a world-class harmonious and livable city, the Beijing Municipal Party Committee and Municipal Government passed the “Opinions on the Implementation of the One-million-mu Afforestation Project in Plain Areas” in 2012. A five-year plan for the afforestation of one million mu in the plain area of Beijing was launched, where in it was decided to build a large-scale urban forest and establish a fully functional urban forest ecosystem with an appropriate layout, resulting in significant benefits.

The implementation of the PAP revolves around a spatial layout consisting of “two rings, three belts, nine wedges, and multiple corridors.” The project focused on greening both sides of the Sixth Ring Road, the sides of key rivers and roads, the aviation corridor and the airport, and important functional areas on both sides of the main line of the South-to-North Water Diversion Project, in addition to demolishing and reestablishing key villages further away from the urban–rural junctions. The project was focused on governance, with a construction scope covering 14 districts and counties, including Shunyi, Tongzhou, Daxing, Fangshan, Changping, Huairou, Pinggu, Mentougou, Yanqing, Miyun and Chaoyang, Shijingshan, Haidian, and Fengtai. Four years later in 2016, 1.05 million mu of afforestation had been completed and more than 54 million trees had been planted. This indicated that the construction task of the million-mu afforestation project in the plain area was successfully achieved and had even exceeded its goals.

This project made an important contribution toward the optimization of urban ecological patterns in Beijing, thereby increasing the ecological well-being of its residents [24,25]. Smooth acceptance of the project combined with the continuous development of follow-up project maintenance work will allow the PAP to play an active role in ecological environment improvement, microclimate adjustment, windbreak and sand fixation, water conservation, air purification, and other ecosystem service capabilities in the areas surrounding Beijing [26,27].

However, to consolidate the achievements of the afforestation project in the Beijing Plain and reach the long-term goal of building an excellent, harmonious, and livable city, a large number of follow-up supporting projects are required. For this purpose, Beijing launched a new round of afforestation and greening construction projects of one million mu in 2018 and issued the “Beijing New Round of One-million-mu Afforestation and Greening Construction Project Master Plan”. The scope of this plan is wider than that of the first project and includes plains, shallow mountainous areas, central urban areas, new towns, and other regions, covering a total area of 10,423 km². According to this plan, the city will add one million mu of forest green space and wetlands by 2022, resulting in the city’s forest coverage rate exceeding 45% and the per capita park green space increasing to 16.6 m².

Since construction was initiated in 2012, the million-mu afforestation project in the plain area of Beijing has completed afforestation of over 130,000 hectares in stages. This project effectively promoted the construction of Beijing’s urban forest system and green network, enhanced the ecological effect of urban green spaces, and changed the ecological woodland landscape of the city. Currently, the city’s forest coverage, forest coverage rate in the plain, urban green coverage, and per capita public green space are approximately 44%, 29.6%, 48.6%, and 16.4 m², respectively. This afforestation project has made history in the Beijing Plain with respect to the construction scale, speed of afforestation, landscape, and overall quality. The implementation of this project also revealed significant challenges for future afforestation and green space management and protection.

3. Materials and Methods

3.1. Data Sources

We conducted a combination of face-to-face interviews and online questionnaire surveys for data collection. Preliminary research was performed to ensure the rationality of the survey; we considered the feasibility of implementation before finalizing the design of the online questionnaire. Preliminary research considered the aspects of project implementation for the creation of parks, green spaces, and green roads; the locations of villages and important gathering places for the population; and population distribution based on interviews with 152 respondents. Data were mainly obtained via the online questionnaire survey. The surveyed residents were located in 16 municipal districts in Beijing, including Mentougou, Yanqing, Huairou, Miyun, Pinggu, Changping, Shunyi, Tongzhou, Daxing, Fangshan, Haidian, Shijingshan, Fengtai, Chaoyang, Xicheng, and Dongcheng districts. The online questionnaire survey was administered to all the resident respondents in Beijing through the Wenjuanxing platform (https://www.wjx.cn/, accessed on 7 February 2023). A total of 1900 questionnaires were distributed, and all of them were recovered. The sampling covered most of the areas of PAP implementation. The questionnaire comprised two main parts: the first part mostly gathered basic information on respondents, such as age, gender, education level, type of work, income level, and length of residency in Beijing; the second part surveyed respondents’ thoughts about the PAP and their satisfaction with the living environment, including aspects such as their knowledge of the PAP and satisfaction with the PAP, landscape effects, ecological function, and recreational spaces. The recovered samples from the respondents were scrutinized to exclude those based on ignorance of the million-mu afforestation project, those that omitted questions, and those that submitted incorrect answers. Finally, 1158 valid questionnaires were retained; this sample size satisfied the requirements of this study.

3.2. Empirical Method

We focused on satisfaction with the PAP as the explanatory variable, which is a categorical variable obtained using the five-point Likert scale. Therefore, the Ologit model of regression was used for analysis [17].

To eliminate the inconsistency and heteroscedasticity of estimation results associated with the Ologit model, it is necessary to convert the observable discrete explained variable, y, into an unobservable continuous variable, y*, via a function. In this study, y* is assumed to be the explained variable’s satisfaction with the PAP. We define:

$$y=\left\{\begin{array}{c}1,y^{*}\le {\theta }_1\\ 2,{\theta }_1<y^{*}\le {\theta }_2\\ 3,{\theta }_2<y^{*}\le {\theta }_3\\ 4,{\theta }_3<y^{*}\le {\theta }_4\\ 5,y^{*}>{\theta }_4\end{array}\right.$$

(1)

Among these, θ₁, θ₂, θ₃, and θ₄ represent unknown cut points, where θ₁ < θ₂ < θ₃ < θ₄. If the latent variable y* is affected by a set of variables, Χ, then y* = α+Χβ+ε, where Χ represents a series of variables affecting employment discrimination, α is a constant term, β represents a parameter vector, and ε represents a random disturbance term that obeys the logistic distribution. The probability of y being 1, 2, 3, 4, and 5 is expressed as:

$$\left\{\begin{array}{c}P\left(y=1\left|x\right.\right)=P\left(y^{*}\le {\theta }_1\left|x\right.\right)\\ \left(y=2\left|x\right.\right)=P\left({\theta }_1\le {y^{*}\le \theta }_2\left|x\right.\right)=P\left(y^{*}\le {\theta }_2\right)-P\left(y^{*}\le {\theta }_1\right)\\ \left(y=3\left|x\right.\right)=P\left({\theta }_2\le {y^{*}\le \theta }_3\left|x\right.\right)=P\left(y^{*}\le {\theta }_3\right)-P\left(y^{*}\le {\theta }_2\right)\\ \left(y=4\left|x\right.\right)=P\left({\theta }_3\le {y^{*}\le \theta }_4\left|x\right.\right)=P\left(y^{*}\le {\theta }_4\right)-P\left(y^{*}\le {\theta }_3\right)\\ \left(y=5\left|x\right.\right)=P\left({y^{*}\ge \theta }_4\left|x\right.\right)=1-P\left(y^{*}\le {\theta }_4\right)\end{array}\right.$$

(2)

Under the logistic distribution, Equation (2) is transformed into:

$$P=\left(Z=z\right)=\mathsf{\Lambda }\left(\mathrm{z}\right)=1/(1+e^{-x})$$

(3)

This leads to:

$$\left\{\begin{array}{c}P\left(y=1\left|x\right.\right)=\mathsf{\Lambda }\left({\theta }_1-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)=\frac{1}{1+e^{\alpha +X\beta +{\theta }_1}}\\ P\left(y=2\left|x\right.\right)=\mathsf{\Lambda }\left({\theta }_2-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)-\mathsf{\Lambda }\left({\theta }_1-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)=\frac{1}{1+e^{\alpha +X\beta +{\theta }_2}}-\frac{1}{1+e^{\alpha +X\beta +{\theta }_1}}\\ P\left(y=3\left|x\right.\right)=\mathsf{\Lambda }\left({\theta }_3-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)-\mathsf{\Lambda }\left({\theta }_2-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)=\frac{1}{1+e^{\alpha +X\beta +{\theta }_3}}-\frac{1}{1+e^{\alpha +X\beta +{\theta }_2}}\\ P\left(y=4\left|x\right.\right)=\mathsf{\Lambda }\left({\theta }_4-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)-\mathsf{\Lambda }\left({\theta }_3-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)=\frac{1}{1+e^{\alpha +X\beta +{\theta }_4}}-\frac{1}{1+e^{\alpha +X\beta +{\theta }_3}}\\ P\left(y=5\left|x\right.\right)=1-\mathsf{\Lambda }\left({\theta }_4-\mathsf{{\rm X}}\beta -\mathsf{\alpha }\right)=1-\frac{1}{1+e^{\alpha +X\beta +{\theta }_4}}\end{array}\right.$$

(4)

Thus, the Ologit regression equation can be obtained from Equation (4) as follows:

$$\ln (\frac{P\left(y\le j\right)}{1-P(y\le j)})={\theta }_j-X\beta -\alpha ,(j=\mathrm{1,2},\mathrm{3,4},5)$$

(5)

It should be noted that the regression coefficients of the Ologit model and their signs do not directly reflect the true degree and direction of the explanatory variables affecting explained variables and can only be used as a basis for the mutual comparison and ranking of explanatory variables [28,29,30,31].

3.3. Variable Selection

The Beijing PAP was initiated and implemented under conditions involving environmental degradation, with particular reference to ecological pressures brought about by smog and extreme weather in recent years. From the perspective of content and goals of the construction, the creation of large-scale, high-level, characteristic, and multi-functional urban forests and adoption of measures that restore wetlands in the plain area will enhance the ecological environment and air quality of Beijing, improve the effects exerted by urban forest landscapes, and increase the well-being of people. Therefore, this project is not limited to simple afforestation. Its implementation may not only confer ecological benefits in the traditional sense but also enhance cultural inheritance, satisfy the green leisure needs of citizens, and enhance the quality of life of the general public by improving the livable environment of the city [20,29]. Considering these criteria, we focused on core explanatory variables that would reflect resident satisfaction and selected the following three variables: (i) landscape effects, (ii) ecological functions, and (iii) recreational spaces [32].

Resident satisfaction with the landscape effects was measured based on the following three aspects: satisfaction with (i) the improvement of local landscapes by afforestation projects, (ii) selection of green seedlings, and (iii) shade afforded by summer trees. We designed the degree of satisfaction according to the five-point Likert scale and requested that respondents choose from numbers 1–5 according to their perception, where 1 = very dissatisfied and 5 = very satisfied.

Satisfaction with ecological functions was measured on the basis of the following three aspects: satisfaction with (i) alleviation of air pollution via afforestation, (ii) noise reduction following afforestation, and (iii) improvement in biodiversity. Similar to that of the previous variable, we designed the degree of satisfaction according to the five-point Likert scale.

Satisfaction with recreational spaces was measured using the following three aspects: satisfaction with (i) improvements made to recreational areas, (ii) accessibility to green spaces, and (iii) overall greening of the environment following afforestation. Similarly, we designed the degree of satisfaction according to the five-point Likert scale and requested that the respondents choose from numbers 1–5, according to their perception.

Other control variables included the individual characteristics of the main residents [8]. These selected control variables included age, gender, education level, occupation type, and salary level. Statistical information pertaining to each variable is listed in

Table 1. Variable statistics.

Variable Category		Indicator	Code	Variable Interpretation
Dependent variable		Satisfaction with the Plain Afforestation Project (PAP)	Y	Overall satisfaction with the afforestation project? (Likert scale *)
Core explanatory variables	Satisfaction with landscape effect	Satisfaction with improved landscape	X1	Has the afforestation project improved the local landscape? (Likert scale)
		Satisfaction with selection of green seedlings	X2	The choice of greening seedlings in afforestation projects? (Likert scale)
		Satisfaction with summer shade	X3	Changes in summer shade after afforestation? (Likert scale)
	Satisfaction with ecological function	Satisfaction with effect of afforestation alleviation of air pollution	X4	Does afforestation reduce air pollution? (Likert scale)
		Satisfaction with noise reduction	X5	Noise reduction after afforestation? (Likert scale)
		Satisfaction with improving biodiversity	X6	Did afforestation projects increase biodiversity? (Likert scale)
	Satisfaction with recreational space	Satisfaction with improving recreational areas	X7	Satisfaction with green recreational areas? (Likert scale)
		Satisfaction with green space accessibility	X8	Satisfaction with green space accessibility? (Likert scale)
		Satisfaction with greening	X9	The overall greening environment satisfaction after afforestation? (Likert scale)
Control variable		Age	X10	Age (1: 7–17; 2: 18–28; 3: 29–44; 4: 45–59; 5: 60+)
		Gender	X11	Gender (0 female; 1 male)
		Profession	X12	Occupational composition of the surveyed population (1 = business unit; 2 = public institutions; 3 = students; 4 = entrepreneurs; 5 = unemployed; 6 = retirees; 7 = freelancers)
		Education	X13	Education level (1 = elementary school; 2 = junior high school; 3 = high school; 4 = undergraduate; 5 = postgraduate or above)
		Salary	X14	Salary level (1 = <5000 yuan; 2 = 5000–10,000 yuan; 3 = >10,000 yuan above)

* Likert Scale: 1 = very dissatisfied; 2 = not very satisfied; 3 = average; 4 = somewhat satisfied; 5 = very satisfied.

.

4. Results and Discussion

4.1. Analysis of Resident Satisfaction

Figure 1 presents an analysis of the degree of resident satisfaction in terms of various functions of the PAP. Overall, the resident satisfaction score of the PAP lies between 3 and 4. Only 14.21% of the residents were very satisfied. The variables for which the largest number of respondents expressed “4 = somewhat satisfied” were improvements in summer shade (39.74%), biodiversity (33.42%), greening level (31.79%), and air quality (24.63%). The variables for which the largest number of respondents expressed “5 = very satisfied” were greening (25.89%) and biodiversity (23.11%). The variable for which the largest number of respondents expressed “3 = average” was alleviation of noise pollution. Resident satisfaction was lowest for improvement in air pollution, with most respondents indicating a score of 3 (average) or less. The remaining variables were similar, receiving a score of 3 (average) or less, with summer shade receiving the lowest negative resident satisfaction (42.48%).

4.2. Factor Analysis of Resident Satisfaction

Prior to the regression analysis, we tested the correlation between variables. The results are presented in

Table 2. Scoring values of each index of landscape quality.

	X1	X2	X3	X4	X5	X6	X7	X8	X9
X1	1.000
X2	0.025	1.000
X3	0.027	−0.016	1.000
X4	0.031	−0.004	−0.010	1.000
X5	−0.001	0.049	−0.051	−0.017	1.000
X6	0.014	0.002	0.026	0.006	0.039	1.000
X7	−0.005	0.037	0.012	−0.006	−0.024	0.002	1.000
X8	0.013	−0.001	0.011	−0.019	−0.025	0.034	0.007	1.000
X9	0.046	−0.035	−0.006	−0.011	0.025	−0.042	0.008	0.021	1.000

. The correlation of each variable (Table 2) was lower than 0.1, which was within a reasonable range and indicated the absence of serious multicollinearity among the explanatory variables.

Regression results of the empirical analysis obtained using the Ologit model are presented in

Table 3. Benchmark regression results.

	(1)	(2)	(3)	(4)
	Landscape Effect	Ecological Function	Recreational Space	All Variables
X1	0.125 ***			0.119 **
	(0.008)			(0.013)
X2	0.128 ***			0.129 ***
	(0.004)			(0.005)
X3	0.141 ***			0.155 ***
	(0.014)			(0.007)
X4		0.131 ***		0.142 ***
		(0.004)		(0.002)
X5		0.118 ***		0.130 ***
		(0.005)		(0.002)
X6		0.110 **		0.116 **
		(0.014)		(0.010)
X7			0.144 ***	0.154 ***
			(0.001)	(0.001)
X8			0.117 **	0.122 ***
			(0.012)	(0.010)
X9			0.106 **	0.118 ***
			(0.012)	(0.006)
X10	0.120 **	0.127 **	0.123 **	0.118 **
	(0.017)	(0.013)	(0.016)	(0.022)
X11	0.276 **	0.240 **	0.240 **	0.262 **
	(0.017)	(0.036)	(0.034)	(0.024)
X12	−0.046 *	−0.047 *	−0.055 **	−0.044 *
	(0.072)	(0.065)	(0.031)	(0.089)
X13	−0.013	−0.002	0.009	0.008
	(0.977)	(0.967)	(0.841)	(0.856)
X14	−0.203 ***	−0.202 ***	−0.180 **	−0.175 **
	(0.008)	(0.009)	(0.020)	(0.025)

* Robust standard errors in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

. The three variables that measured satisfaction with landscape effects exerted a positive impact on satisfaction with the PAP at the 1% significance level (Table 3, column 1). A comparison of the regression coefficients of the three variables indicated that satisfaction with shade in summer exerted the greatest influence on satisfaction with the PAP, followed by satisfaction with the selection of greening seedlings and satisfaction with landscape improvement. Increased shade in summer is associated with more lush urban forests and higher canopy closure [33]. Residents had an intuitive understanding of the shade afforded by trees in summer, which specifically reflected the efficacy of the PAP’s implementation.

Resident satisfaction with forest-based alleviation of air pollution and noise had a positive impact on satisfaction with the PAP at the 1% significance level (Table 3, column 2). It also had a significant positive effect on improving biodiversity satisfaction at a significance level of 5%. In contrast, satisfaction with forest air pollution mitigation showed the greatest impact on satisfaction with the PAP, followed by satisfaction with noise mitigation and biodiversity improvement. Recently, extreme disasters, such as dust and smog, have been occurring frequently, forcing residents to demand better air quality [4]. Therefore, the higher the satisfaction with the mitigation of air pollution due to afforestation, the higher the satisfaction with the PAP.

Satisfaction with the improvement of recreational areas had a positive impact on the PAP at the 1% significance level (Table 3, column 3). Satisfaction with green space accessibility and greening also had a positive impact on satisfaction with the PAP at the significance level of 5%. A comparison between coefficients showed that among the three variables measuring satisfaction with recreational space, satisfaction with the improvement of recreational spaces had the strongest promoting effect on satisfaction with the PAP, followed by satisfaction with accessibility to green spaces and greening.

Each variable measuring satisfaction with landscape effects, ecological functions, and recreational spaces had a significant positive impact on satisfaction with the PAP (Table 3, column 4). These results are more consistent with those of the first three columns. A comparison of the coefficients of the core explanatory variables showed that the top three criteria related to satisfaction with the PAP were satisfaction with summer shade, improvement of recreational areas, and air pollution alleviation.

Results of the control variables showed that age exerted a significant positive effect on satisfaction with the PAP at the 1% significance level (Table 3, column 4). This indicates that older residents were more easily satisfied with the PAP. This may be attributed to the fact that older residents, having personally experienced Beijing’s environment gradually changing from poor to good, were more likely to be satisfied with the PAP [24].

Male residents were more likely than female residents to be satisfied with the PAP at the 5% significance level. A possible reason is that women may be more sensitive to the external environment than men and thus have higher expectations from the PAP [34,35]. Cheung et al. [32] also reported a similar conclusion.

The higher the education level, the lower the satisfaction with the PAP at the 10% significance level. A possible reason is that residents with a high level of education have higher expectations for urban living environments, greening, and recreation and are therefore less satisfied with the PAP [36].

Residents with high income levels were also less satisfied with the PAP. Generally, the higher the income, the higher the environmental quality requirements and the more difficult it is to be satisfied with the PAP [37].

4.3. Robustness Test

To test the robustness of our empirical analysis, we conducted a robustness analysis by transforming the regression method and randomly selecting regression samples [38,39]. Results of the robustness test are presented in

Table 4. Results of the robustness test.

	(1)	(2)	(3)
Y	OLS	Poisson	Randomly drawn regression samples
X1	0.072 **	0.022 ***	0.031 ***
	(0.011)	(0.011)	(0.008)
X2	0.075 ***	0.023 ***	0.034 ***
	(0.007)	(0.007)	(0.005)
X3	0.093 ***	0.028 ***	0.039 ***
	(0.006)	(0.006)	(0.003)
X4	0.085 ***	0.026 ***	0.022 ***
	(0.002)	(0.002)	(0.049)
X5	0.070 ***	0.021 ***	0.021 ***
	(0.005)	(0.005)	(0.041)
X6	0.073 ***	0.022 ***	0.020 ***
	(0.006)	(0.006)	(0.067)
X7	0.090 ***	0.027 ***	0.023 ***
	(0.001)	(0.001)	(0.042)
X8	0.080 ***	0.024 ***	0.035 ***
	(0.005)	(0.005)	(0.002)
X9	0.070 ***	0.021 ***	0.023 ***
	(0.006)	(0.007)	(0.025)
X10	0.076 **	0.023 **	0.026 **
	(0.012)	(0.011)	(0.033)
X11	0.120 *	0.037 *	0.070 ***
	(0.071)	(0.062)	(0.008)
X12	−0.027 *	−0.008*	−0.006
	(0.077)	(0.080)	(0.303)
X13	0.002	0.001	−0.003
	(0.940)	(0.932)	(0.755)
X14	−0.084 *	−0.025 *	−0.017
	(0.065)	(0.067)	(0.349)
_cons.	0.980 ***	0.489 ***	0.345 ***
	(0.004)	(0.000)	(0.016)

* Robust standard errors in parentheses: *** p < 0.01, ** p < 0.05, * p <0.1.

. In Table 4, the first, second, and third columns show the regression results of ordinary least squares (OLS), Poisson, and randomly selected regression samples, respectively. Regardless of whether OLS or Poisson was used, the significance of the regression coefficients of the core variables was consistent with the results of the Ologit regression. Similarly, when 1,000 samples were randomly selected, the regression results of the core variables were not significantly different from those of previous results. These test results show that the regression results were relatively robust.

5. Conclusions and Suggestions

In this study, we utilized an Ologit model in conjunction with 1,158 survey samples from Beijing to empirically analyze resident satisfaction with the PAP. The results showed that improving satisfaction with the landscape effect following afforestation, as well as satisfaction with ecological functions and recreational spaces, may significantly improve the overall resident satisfaction with the afforestation project. Specifically, the key factors affecting resident satisfaction with PAPs may be ranked in order of contribution as follows: satisfaction with (i) shade in summer; (ii) improvement of recreational areas; (iii) air pollution mitigation; (iv) noise reduction and selection of green seedlings; (v) green space accessibility; (vi) improved landscape; (vii) greening; and (viii) improved biodiversity. Furthermore, the results indicated that older residents are more likely than younger residents to be satisfied with the PAP, and residents with higher levels of education and income tend to be less satisfied with the PAP.

Based on our results, we propose the following policy initiatives:

(1)

Improve green landscapes. To this end, the Beijing government must first continue to increase the total amount of forest resources and then launch green space landscaping, improve the landscape of newly planted forest land, and focus on connectivity between different landscapes. Finally, attention should be focused on building a regional landscape suited to Beijing’s specific characteristics, thereby facilitating a sense of belonging in residents.

(2)

Create a sustainable ecological environment system and promote the continuous development of the various ecological functions of forest resources. The importance of multi-functional collocation of different tree species during the afforestation process should emphasize the importance of cultivating near-natural forests. The Beijing government should provide funding for the implementation of ecosystem management, biodiversity protection, and near-nature management, enabling the growth of a suburban forest with independent ecosystem development functions and improving the level of forest summer shade to maximize the ecological service value of the PAP.

(3)

Improve supporting facilities, provide leisure parks, and increase investment in green spaces; support public facilities; improve accessibility to green spaces; respond to the needs of leisure markets; and enrich the spiritual life of residents. Moreover, the Beijing government should establish a social feedback mechanism via the internet and provide satisfaction-based questionnaires on government information disclosure websites or other specialized websites to regularly collect social feedback related to the PAP and adjust policy guidelines accordingly.

As forest city construction in cities around the world continues, large-scale projects aimed at greening and afforestation have begun. This research was performed to deepen understanding of the satisfaction of residents with afforestation projects based on its improvement to their livelihoods and other major interests. Although the indexing system used in this study focused on Beijing, it is necessary to establish a unified and detailed indexing system that reflects resident satisfaction with PAPs in the future to ensure that standard evaluation and comparative analysis of afforestation effects can be conducted nationwide. Moreover, since resident satisfaction with afforestation effects varies with time, space, and the level of respondent awareness, it may be necessary in the future to analyze the spatial distribution of satisfaction with afforestation projects in the plains of Beijing and continue tracking resident satisfaction for 5 years or more after afforestation.