Promoting the Tripartite Cooperative Mechanism of E-Commerce Poverty Alleviation: Based on the Evolutionary Game Method

Abstract

E-commerce agriculture has gradually become an important force in poverty alleviation. Some large e-commerce enterprises in China, such as Alibaba and JD, have carried out poverty alleviation activities with the support of the government. With the enthusiasm of consumers towards the goal of helping farmers, they can continue to develop e-commerce agriculture and efficiently achieve this goal. However, the unstable relationship among e-commerce enterprises, the government, and consumers with regard to poverty alleviation hinders the realization of poverty reduction goals. To promote the cooperation among the three parties under e-commerce poverty alleviation, this study uses the evolutionary game method. First, this study analyzes the collaborative mechanism of the government, e-commerce enterprises, and consumers participating in poverty alleviation. Second, based on the above analysis, a tripartite evolutionary game model is constructed. A payment matrix is established to analyze the factors that affect the strategic choices of participants, and the conditions for promoting the tripartite collaborative mechanism of poverty alleviation are discussed using numerical simulation. The results show the following: (1) The reduction in poverty alleviation costs and the increase in cooperation benefits are important factors in promoting the choice of poverty alleviation strategies. (2) E-commerce enterprises are more sensitive to consumer preferences for helping farmers. When consumer preferences for helping farmers are low, the government’s early efforts to participate in poverty alleviation can guide e-commerce enterprises and consumers to participate in cooperative poverty alleviation. (3) The government and e-commerce enterprises can reduce poverty alleviation incentives and agricultural product premiums by increasing the willingness of e-commerce enterprises and consumers to support farmers in the initial stage to promote the participation of all parties in poverty alleviation at a relatively low cost.

1. Introduction

With the rapid development of e-commerce agriculture, e-commerce enterprises have gradually become an important force in helping poor farmers [1]. To solve the problem of poverty, the government has tried to reduce poverty for poor farmers by calling on e-commerce enterprises to participate in poverty alleviation [2,3]. In the Guidance on Promoting E-commerce Targeted Poverty Alleviation issued by the government, it was also proposed that e-commerce enterprises “Actively encourage major e-commerce enterprises and social forces to participate in the fight against poverty, and drive farmers in poor areas out of poverty” [4]. Many types of e-commerce enterprises, such as platform e-commerce (Alibaba) and short video e-commerce (Tiktok), have carried out agricultural assistance with the support of the government. This means that the government will cooperate with e-commerce enterprises and consumers to participate in poverty alleviation activities. Recent studies provide evidence that with the participation of these three parties in poverty alleviation activities, we can improve the scale and quality of agricultural production in poor areas, open the market with marketable agricultural products and services to achieve a virtuous cycle of supply and demand, and provide long-term and stable internal support for poor farmers to alleviate poverty [5,6,7]. It can be seen that the development of e-commerce poverty alleviation activities cannot be separated from the need for joint participation and cooperation.

However, in the process of implementing and promoting e-commerce poverty alleviation, the unstable cooperative relationship among the three parties gradually becomes obvious [8,9,10]. The government’s mode of promoting poverty alleviation is not mature, and its guiding role in e-commerce poverty alleviation has not been fully played [11,12]. The overall efficiency of poverty alleviation efforts by e-commerce enterprises is not high [13]. To balance the costs, risks, and benefits of agricultural aid investment, e-commerce enterprises do not take poverty alleviation as a strategic priority [14], and the proportion and motivation of consumers that are willing to help farmers are not strong enough [15]. These problems lead to a failure to reach the consensus of members to participate in poverty alleviation through collaboration, which hinders the completion of poverty alleviation goals.

Much research has been conducted on issues such as promoting e-commerce poverty alleviation, but we found that most studies describe one-way impacts between two poverty alleviation subjects. Some scholars propose that e-commerce enterprises can be encouraged to actively participate through government incentives [16,17], and some scholars believe that the efforts of e-commerce enterprises and the government can stimulate consumers’ enthusiasm for poverty alleviation [18,19]. In practice, the attitude of consumers towards poverty alleviation will in turn affect the decisions of other members involved in poverty alleviation, and the decisions of members involved in collaborative poverty alleviation will also be affected by the collaborative poverty alleviation mechanism.

Based on the above questions, this paper analyzes the collaborative poverty alleviation mechanism by which the government, e-commerce enterprises, and consumers participate in poverty alleviation, focusing on the measurement of their own benefits by the members involved in poverty alleviation, and examines the synergistic effect of the tripartite poverty alleviation behaviors. Then, through the evolutionary game theory, an evolutionary game model of tripartite cooperative agricultural assistance consisting of the government, e-commerce enterprises, and consumers is constructed. This paper also simulates relevant models, analyzes the main factors affecting the decision-making process of all participants from the perspective of the interests of the three parties, and puts forward suggestions and countermeasures to guide the promotion of a sustainable e-commerce coordinated poverty alleviation mechanism.

Our research contributions are mainly reflected in three aspects. First, consumers are included as the decision-making party, and the impact of psychological factors such as consumers’ poverty alleviation preference and utility in e-commerce poverty alleviation is analyzed. Second, we consider the collaborative effect of the government, e-commerce enterprises, and consumers on each other in the process of poverty alleviation. Finally, evolutionary game theory is applied to study the influence of different factors of e-commerce poverty alleviation on each party’s choice of strategy, providing a reference for the promotion of cooperation in e-commerce poverty alleviation.

The rest of the article is arranged as follows. Section 2 provides a literature review. In Section 3, we analyze and construct a model. This part assumes the parameters involved in e-commerce anti-poverty cooperation, and constructs a tripartite evolutionary game model of the government, e-commerce enterprises, and consumers. Section 4 discusses the strategy choices of the participants and the balance of the model. In Section 5, the simulation analysis is carried out, and the conditions for promoting tripartite participation in poverty alleviation are discussed. Section 6 provides our conclusions and opinions.

2. Literature Review

2.1. Poverty Alleviation Research concerning the Government, E-Commerce Enterprises, and Consumers

The research on poverty alleviation mainly focuses on governance. In the study of government poverty alleviation, some scholars believe that government investment in agricultural technology can improve the production and income of agricultural products [20], and the combination of policy guidance and support for farmers’ development conditions is better able to achieve poverty alleviation [21,22]. In the research on e-commerce enterprises’ poverty alleviation efforts, some scholars believe that e-commerce enterprises are gradually becoming an effective force for helping farmers [23], and that the e-commerce model has a positive impact on consumers’ willingness to buy [24]. Additionally, e-commerce enterprises can help farmers by increasing farmers’ income and encouraging multi-dimensional poverty alleviation [6]. In addition, the government can use e-commerce to improve the welfare of farmers [2,3]. The government can stimulate consumers’ enthusiasm to help farmers by subsidizing and supporting e-commerce enterprises, and can promote the operation of an e-commerce poverty alleviation model [16,17,25]. As for the study of consumers, Moon [26] investigated the attitude of American citizens towards government support for agriculture, and found that consumers hold a supportive attitude. Block [27] believes that the attitude of consumers, their experience of products and services, and the transformation of their consumption capacity are conducive to alleviating poverty. Huang [28] believes that when the government carries out poverty alleviation, it not only needs economic support but should also increase consumers’ positive psychological feelings. Zhou [18] believes that consumers have a preference for helping farmers, and proposes that enterprises improve consumers’ willingness to buy through agricultural advertising, and obtain a positive public image from it.

The above studies show that the government, e-commerce enterprises, and consumers have a positive impact on poverty alleviation, and there is a certain cooperation mechanism. However, few studies have considered the effect of the three parties’ coordinated poverty alleviation. However, these studies provide a theoretical basis for building a coordinated poverty alleviation model composed of the government, e-commerce enterprises, and consumers, guiding participants to actively help the poor.

2.2. Research on E-Commerce Poverty Alleviation Cooperation

Although the current e-commerce poverty alleviation strategy has generally achieved positive results, it is still necessary to strengthen market supervision and the self-discipline of the platform, and encourage all social entities to participate in poverty alleviation [29]. By describing the implementation of the rural poverty alleviation project in Kalesang Village, Nawawi [30] found that the coordination and cooperation between the government and the community can easily solve the village poverty problem. McKague [11] provided a poverty alleviation framework of comprehensive cooperation, emphasizing the different roles of the private sector, government, and civil society organizations in poverty reduction. However, there are many problems in the e-commerce anti-poverty cooperation model, such as the low participation of all subjects, insufficient resource integration, poor sustainability, and difficulty in balancing interests, rights, and responsibilities [19,31]. Muslims [8] linked collaborative governance with poverty alleviation through systematic map research, and found that the proportion of collaborative governance applied to poverty alleviation was the lowest. Larantika [9] found through interviews that the cooperation between various units in poverty alleviation is limited.

As for how to promote the participation of participants in cooperation, Pant and Laxmi Prasad [32] proposed that all social subjects can be guided to cooperate through morality and a sense of responsibility to achieve their social value. Xu [33] believed that government supervision and the active participation of all parties are necessary conditions for achieving coordinated governance, and the government should play a leading role in the process of poverty alleviation [21]. Teng [34] found that effective government supervision can enable farmers to carry out green production and encourage consumers to buy high-quality agricultural products, and jointly develop the green agricultural product market.

The previous research has laid a foundation for promoting coordinated poverty alleviation. However, although the government’s supervision responsibility is often taken into consideration, the synergy in the process of e-commerce poverty alleviation has been less studied. Practice shows that the sustainable development of e-commerce poverty alleviation also depends on the participation of e-commerce enterprises and the support of consumers.

2.3. Research on Cooperation Mechanism Based on Game Theory

Game theory is often used to solve coordination problems. Liu [35] built a game of emergency rescue cooperation among government rescue teams, social organizations, and government support institutions, and pointed out that the trust relationship and institutional norms among game subjects can promote tripartite cooperation. In order to promote the joint operation of hydropower stations to improve the utilization rate of water resources, Chen [36] built a tripartite game model consisting of two upstream hydropower stations and one downstream hydropower station, and found that the choice of strategy of the game subject depends on the behavior of the partners and their own net income. The above shows the effective cooperation mechanism, and the influence of the behavior and income of members on the cooperation problem, but it does not combine with the problem of agricultural poverty alleviation.

Wan [37] used a poverty alleviation game model including cooperatives, smart supply chain platforms, and the government, and concluded that consumer preferences not only affect the cooperative game between the smart supply chain and partners but also affect product demand and product prices. The regulatory measures of the leading party help to promote the cooperation of the participants. Huo [38] built a game model of farmers’ cooperatives, manufacturers, and retailers to solve the problem of unstable cooperation in the supply chain of agricultural products. Research shows that strict supervision of farmers’ cooperatives promotes coordination and cooperation between all parties. Ju F [39] built a tripartite evolution model in the ecological compensation poverty alleviation project, and found that government regulation depends on the expected benefit cost. A reasonable reward and punishment system also has a positive impact on promoting cooperation. Gou [40] found that by building a tripartite evolutionary game model of agricultural service providers, farmers, and the government, the government can encourage farmers to actively produce by improving reward and punishment and strengthening supervision. Zhao and others [41] established a game model of tripartite collaborative governance between the government, third-party governance, and farmers, and found that cost–benefit, reward, and punishment mechanisms jointly affected the strategies of all parties. Zhang N [42] found that the punishment mechanism can constrain the strategic choice of social organizations and poor groups by building a tripartite evolutionary game between the government, social organizations, and poor groups in the process of poverty alleviation, but the increase in group interests increases the willingness of both parties to cooperate. The above research shows that the application of an evolutionary game model can effectively solve the cooperation problem of participants. However, few studies have considered the impact of consumer behavior on poverty alleviation members in the cooperative poverty alleviation mechanism. It is undeniable that the government’s guidance, the reward and punishment mechanism, and the cost of cooperation are important factors affecting the cooperative decision-making process of members.

3. Materials and Methods

3.1. Description of the Tripartite Game Model

The e-commerce poverty alleviation model injects funds and technologies into poor areas by mobilizing the power of the government and e-commerce enterprises. Relying on the information advantages of e-commerce enterprises to fully understand market demand, poor areas are guided to produce and provide high-quality agricultural products that satisfy consumers [5]. The consumption demand of consumers will stimulate the production capacity of rural areas, help rural areas to increase income, and provide healthy and stable internal support for poverty eradication [4]. The government, e-commerce enterprises, and consumers, as the main participants in e-commerce poverty alleviation, play different roles in the activities of poverty alleviation, and the mechanisms leading to the decisions made by each subject are also different. As the government, e-commerce enterprises, and consumers play different roles in poverty alleviation activities, and the decision-making mechanisms leading to each subject are also different, it is often difficult for the three parties to reach a consensus on collaborative poverty alleviation, which hinders the sustainable and healthy development of the e-commerce poverty alleviation system. Therefore, it is necessary to distinguish the role of the participants in the process of poverty alleviation and the important factors for them to measure their own benefits, reduce the poverty alleviation dilemma caused by the contradiction of interests of the participants, and truly play a role in the poverty alleviation coordination mechanism.

The government-led poverty alleviation activities mostly focus on the improvement of farmers’ production conditions, technologies, and transportation conditions [43,44]. The fundamental goal is to improve the rural economy and the quality of agricultural products. The improvement of the farmers’ economy and the quality of agricultural products brought about by poverty alleviation are the primary considerations of the government.

The poverty alleviation led by e-commerce enterprises can not only help guide farmers to plant scientifically from the production side [45] but also drive the growth of the agricultural product consumption market through the platform flow advantage [24]. As enterprises with rational economic thinking, e-commerce enterprises should not only respond to the government’s agricultural policy but also safeguard their own interests. E-commerce enterprises take the maximization of their own interests as the starting point to measure the cost and income of poverty alleviation.

Consumers drive the economy of agricultural products by purchasing agricultural products. They are mainly faced with the psychological utility brought about by the behavior of helping farmers and the quality of agricultural products purchased. Consumers hope to buy relatively high-quality agricultural products [27] while supporting farmers and agriculture [46], so they make decisions based on the principle of maximizing their own utility. Through the above analysis of the poverty alleviation gains of all parties, it can be seen that the poverty alleviation of any party will not only support farmers but also produce positive benefits for other participants.

Here, we analyze the coordinated poverty alleviation mechanism composed of the government, e-commerce enterprises, and consumers. When the three parties reached a consensus on coordinated poverty alleviation, e-commerce enterprises and the government mainly focused on improving the production quality of agricultural products. E-commerce enterprises not only cooperated with the government to help farmers from the supply side but also cooperated with consumers to promote the increase in agricultural product income from the demand side. At this time, the interaction of the three parties not only played a role in poverty alleviation but also had positive benefits for the members, promoting their sustained, coordinated poverty alleviation. The collaborative mechanism of the e-commerce poverty alleviation system is shown in Figure 1.

3.2. Construction of the Tripartite Game Model

Through the description and analysis of the tripartite cooperative poverty alleviation model, a cooperative poverty alleviation model including the government, e-commerce enterprises, and consumers is constructed. Some assumptions about the evolutionary game model are as follows:

Assumption 1.

In view of the fact that poverty alleviation is a national policy, the government will choose to support farmers and agriculture, but the strength of poverty alleviation is different. The government has two strategies: “vigorously” and “not vigorously”. The probability of vigorously implemented poverty alleviation is x. E-commerce enterprises choose to participate in or not participate in poverty alleviation according to the principle of maximizing benefits; the probability of “participation” is y, and the probability of “no participation” is 1-y. Consumers have two strategies for poverty alleviation: “helping” and “no helping”. The probability of consumers helping farmers is z. All three parties are determined to be rational.

Assumption 2.

As for the impact of the tripartite poverty alleviation strategy on farmers, assuming that the quality of the original agricultural products is$G$, when there is only vigorous assistance from the government, the quality of farmers’ agricultural products is improved;$G_1$. When only e-commerce enterprises participate in poverty alleviation activities, the quality of agricultural products is improved;$G_2$. When the government makes great efforts to support agriculture and e-commerce enterprises participate in helping farmers, the quality of agricultural products is improved;$G_3$. As the government and e-commerce enterprises work together to support agriculture, they have a cooperative effect, which can help farmers to produce and transport agricultural products more comprehensively and effectively;$G_3>(G_1+G_2)$. At the same time, the farmers’ economic development can be driven by the help of e-commerce enterprises and consumers. Assuming that both parties support farming at the same time, the income of farmers from selling agricultural products is${\lambda }_{2}V$. When either party helps farmers, farmers’ income is${\lambda }_{1}V$. When neither of the two parties help farmers, the farmers’ income is$V$($\lambda$is the yield coefficient of poverty alleviation and$1<{\lambda }_1<{\lambda }_2$).

Assumption 3.

The cost of the government vigorously implementing poverty alleviation is$C_g$, and the reward for e-commerce enterprises participating in agricultural aid is$T$. When the government provides loosely implemented poverty alleviation, the government’s payment level is expressed as a decrease in costs and incentives. At this time, the cost of easing the implementation of poverty alleviation is$\eta C_g$, where$\eta$is the extent of the government’s implementation of poverty alleviation activities$(0<\eta <1)$, and the amount of incentives for e-commerce enterprises has also been reduced to$\eta T$. The government’s income is reflected by measuring the overall effect of poverty alleviation. The higher the income of farmers and the better the quality of agricultural products, the higher the degree of completion of the government’s goal of poverty alleviation. Quantifying the government’s return as the parameters related to farmers’ income and agricultural product quality, we obtain${\alpha }_1{\lambda }_{i}V$and${\alpha }_2(G+G_j)$(i = 1, 2; j = 1, 2, 3) (α₁ represents the government’s measurement coefficient of farmers’ economy, and α₂ represents the government’s measurement coefficient of agricultural product quality).

Assumption 4.

The income of e-commerce enterprises through the sale of agricultural products is $\theta {\lambda }_{i}V(i=1,2)$ (assuming that the income of e-commerce enterprises is drawn from the income of farmers according to the commission ratio of $\theta$), and the cost of e-commerce enterprises participating in poverty alleviation is $C_e$. At this time, the reduction ratio of the commission charged to farmers is $\mu$.

Assumption 5.

When consumers choose not to help farmers, the price of agricultural products is $P$, and when consumers choose to help farmers, they are willing to pay a certain agricultural aid premium, $\Delta P$, for agricultural products [47]. Consumers will generate psychological satisfaction benefits to help farmers, assuming that the psychological satisfaction benefit that consumers receive from helping farmers is $f_1\Delta U$ ($f_1$ is the consumers’ preference for helping farmers, and $\Delta U$ is the consumers’ basic benefit for helping farmers). The quality of the agricultural products received affects the utility of consumers at the same time. The utility of consumers with regard to the quality of agricultural products is $f_2(G+G_i)(i=1,2,3)$ ($f_2$ refers to the consumers’ preference for the quality effect of agricultural products). When consumers actively participate in helping farmers, they will be appropriately tolerant of agricultural product quality problems, the utility of agricultural product quality increases by $b$ times, and they have a positive evaluation of the government and e-commerce enterprises participating in poverty alleviation. The reputation of the government and e-commerce enterprises increases; $R_g$, $R_e$.

The parameters mentioned in the assumptions and their meanings are shown in

Table 1. Related parameters and their meanings.

Parameters	Meanings
$G$, $V$	Initial agricultural product quality and sales income
$G_1$$,G_2$$,G_3$	Increase in agricultural product quality when the government vigorously helps the poor, e-commerce enterprises participate in poverty alleviation, and the two cooperate in poverty alleviation
${\lambda }_{1}V$$,{\lambda }_{2}V$	Income from agricultural product sales when either e-commerce enterprises or consumers help the poor, or both help the poor
$C_g$$,\eta C_g$	The cost when the government chooses to help the poor vigorously or not
$T$$,\eta T$	Reward value of e-commerce enterprises participating in poverty alleviation when the government is vigorously or not vigorously helping farmers
${\alpha }_1$$,{\alpha }_2$	The government’s measurement value coefficient of agricultural economy and agricultural product quality
$C_e$	Cost of e-commerce enterprises participating in poverty alleviation
$\theta$	Commission rate drawn by e-commerce enterprises for agricultural product sales
$\mu$	Reduction ratio of commission rate when e-commerce enterprises participate in poverty alleviation
$P$$,\Delta P$	The price of agricultural products and the premium that consumers are willing to pay when helping farmers
$f_1$	Consumers’ preference for helping farmers
$\Delta U$	Basic utility value of consumers helping farmers
$f_2$	Utility measurement coefficient of consumers on agricultural product quality
$b$	The multiple of increase in the utility value of agricultural product quality when consumers help farmers
$R_g$$,R_e$	Reputation added value when the government vigorously helps the poor and e-commerce enterprises participate in poverty alleviation

.

3.3. Building the Tripartite Payment Matrix

The effect of e-commerce poverty alleviation depends on the results of the dynamic game between the government, e-commerce enterprises, and consumers. On the basis of the above model assumptions, combined with the strategic choices of the government, e-commerce enterprises, and consumers, a tripartite payment matrix is constructed, as shown in

Table 2. Payment matrix of a tripartite evolutionary game.

Government	E-Commerce Enterprises	Consumers
		Helping (z)	No Helping (1 − z)
Vigorously (x)	Participation (y)	${\alpha }_1{\lambda }_{2}V+{\alpha }_2(G+G_3)-C_g-T+R_g$ $(1-\mu )\theta {\lambda }_{2}V-C_e+T+R_e$ $f_1\Delta U+b{f}_2(G+G_3)-P-\Delta P$	${\alpha }_1{\lambda }_{1}V+{\alpha }_2(G+G_3)-C_g-T$ $(1-\mu )\theta {\lambda }_{1}V-C_e+T$ $f_2(G+G_3)-P$
	No participation $(1-\mathrm{y})$	${\alpha }_1{\lambda }_{1}V+{\alpha }_2(G+G_1)-C_g+R_g$ $\theta {\lambda }_{1}V$ $f_1\Delta U+b{f}_2(G+G_1)-P-\Delta P$	${\alpha }_{1}V+{\alpha }_2(G+G_1)-C_g$ $\theta V$ $f_2(G+G_1)-P$
Not vigorously $(1-\mathrm{x}$)	Participation (y)	${\alpha }_1{\lambda }_{2}V+{\alpha }_2(G+G_2)-\eta C_g-\eta T$ $(1-\mu )\theta {\lambda }_{2}V-C_e+\eta T+R_e$ $f_1\Delta U+b{f}_2(G+G_2)-P-\Delta P$	${\alpha }_1{\lambda }_{1}V+{\alpha }_2(G+G_2)-\eta C_g-\eta T$ $(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T$ $f_2(G+G_2)-P$
	No participation $(1-\mathrm{y})$	${\alpha }_1{\lambda }_{1}V+{\alpha }_{2}G-\eta C_g$ $\theta {\lambda }_{1}V$ $f_1\Delta U+b{f}_{2}G-P-\Delta P$	${\alpha }_1{\lambda }_{2}V+{\alpha }_2(G+G_3)-C_g-T+R_g$ $\theta V$ $f_{2}G-P$

Note: From top to bottom are the benefits of government, e-commerce enterprises, and consumers.

.

4. Analysis of the Evolutionary Stability Strategy of Three Parties

4.1. Analysis of the Government Evolution

The expected payoff of the government when it chooses the “vigorously” strategy is:

$$\begin{array}{l}{E}_{g1}=yz{\alpha }_1{\lambda }_{2}V+y(1-z){\alpha }_1{\lambda }_{1}V+y[{\alpha }_2(G+G_3)-C_g-T]+z{R}_g+(1-y)z{\alpha }_1{\lambda }_{1}V\\ +(1-y)(1-z){\alpha }_{1}V+(1-y)[{\alpha }_2(G+G_1)-C_g]\end{array}$$

(1)

Additionally, the expected payoff of the government when it chooses the “not vigorously” strategy is:

$$\begin{array}{l}{E}_{g2}=yz{\alpha }_1{\lambda }_{2}V+y(1-z){\alpha }_1{\lambda }_{1}V+y[{\alpha }_2(G+G_3)-\eta C_g-\eta T]+(1-y)z{\alpha }_1{\lambda }_{1}V\\ +(1-y)(1-z){\alpha }_{1}V+(1-y)[{\alpha }_2(G+G_1)-\eta C_g]\end{array}$$

(2)

Thus, the average payoff of the government is:

$${\overline{E}}_g=x{E}_{g1}+(1-x)E_{g2}$$

(3)

It can be seen from the above formula that the government’s replication dynamic equation for strategy selection is:

$$\begin{array}{l}F(x)=dx/dt=x(E_{g1}-{\overline{E}}_g)\\ =x(1-x)\{z{R}_g+{\alpha }_2{G}_1-(1-\eta )C_g+y[{\alpha }_2(G_3-G_2-G_1)-(1-\eta )T]\}\end{array}$$

(4)

Let $F(x)=0$; there is: $x=0$, $x=1$, $y=y^{*}=\frac{(1-\eta )C_g-z{R}_g-{\alpha }_2{G}_1}{{\alpha }_2(G_3-G_2-G_1)-(1-\eta )T}$.

If $y=y^{*}$, no matter what proportion of the “vigorously” strategy is chosen, the government’s strategic choice is always in a stable state. If $y<y^{*}$, there are $\frac{dF(x)}{dx}{|}_{x=0}<0$ and $\frac{dF(x)}{dx}{|}_{x=1}>0$, then $x=0$ is ESS. That is, the government will choose the “not vigorously” strategy. If $y>y^{*}$, there are $\frac{dF(x)}{dx}{|}_{x=0}>0$ and $\frac{dF(x)}{dx}{|}_{x=1}<0$, then $x=1$ is ESS. That is, the government will choose the “vigorously” strategy.

The dynamic evolutionary process of the government’s strategic choice is shown in Figure 2. We can draw some simple relationships from Figure 2b,c: The government’s choice of strategy is affected by the strategies of e-commerce enterprises and consumers. When the government opts for the “not vigorously” choice of poverty alleviation, that is, in the state of $y<y^{*}$, if the total revenue of the government’s vigorously implemented poverty alleviation increases and the total cost is less at this time, improving the quality of agricultural products in collaboration with e-commerce enterprises, increasing the social reputation generated by vigorously supported farmers and agriculture, and reducing the incentives for e-commerce enterprises to participate in helping farmers, this will cause $y=y^{*}=\frac{(1-\eta )C_g-z{R}_g-{\alpha }_2{G}_1}{{\alpha }_2(G_3-G_2-G_1)-(1-\eta )T}$ to decrease, and enter a state of $y>y^{*}$. Then, the government will change their strategy, becoming more inclined to vigorously supporting the agricultural strategy. When the growth of agricultural products quality caused by the e-commerce platform’s poverty alleviation strategy is relatively high, there is $y<y^{*}$; the government has the motivation of “hitchhiking”, and will choose the “vigorously” strategy, that is, it will tend to relax support for farmers and agriculture.

4.2. Analysis of the E-Commerce Enterprises’ Evolution

The expected payoff of e-commerce enterprises when they choose the “participation” strategy is:

$$\begin{array}{l}{E}_{e1}=xz[(1-\mu )\theta {\lambda }_{2}V-C_e+T+R_e]+x(1-z)[(1-\mu )\theta {\lambda }_{1}V-C_e+T]\\ +(1-x)z[(1-\mu )\theta {\lambda }_{2}V-C_e+\eta T+R_e]+(1-x)(1-z)[(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T]\end{array}$$

(5)

The expected payoff of e-commerce enterprises when they choose the “no participation” strategy is:

$$E_{e2}=z\theta {\lambda }_{1}V+(1-z)\theta V$$

(6)

Therefore, the average payoff of e-commerce enterprises is calculated as:

$${\overline{E}}_e=y{E}_{e1}+(1-y)E_{e2}$$

(7)

The corresponding replicator dynamic equation for e-commerce enterprises’ strategy selection is:

$$\begin{array}{l}F(y)=dy/dt=y(E_{e1}-{\overline{E}}_e)\\ =y(1-y)\{\eta T+x(T-\eta T)-C_e+(1-\mu )\theta {\lambda }_{1}V-\theta V+z[(1-\mu )\theta ({\lambda }_2-{\lambda }_1)V+R_e-\theta {\lambda }_{1}V+\theta V]\}\end{array}$$

(8)

Let $F(y)=0$; there is: $y=0$, $y=1$, $z=z^{*}=\frac{\theta V-\eta T-x(T-\eta T)+C_e-(1-\mu )\theta {\lambda }_{1}V}{(1-\mu )\theta ({\lambda }_2-{\lambda }_1)V+R_e-\theta {\lambda }_{1}V+\theta V}$.

If $z=z^{*}$, regardless of the probability that e-commerce enterprises choose to participate in poverty alleviation, their strategic choices are always stable. If $z<z^{*}$, we have $\frac{dF(y)}{dy}{|}_{y=0}<0$ and $\frac{dF(y)}{dy}{|}_{y=1}>0$, and thus $y=0$ is the ESS; e-commerce enterprises will choose not to participate in the poverty alleviation strategy. If $z>z^{*}$, then $\frac{dF(y)}{dy}{|}_{y=0}>0$ and $\frac{dF(y)}{dy}{|}_{y=1}<0$, implying $y=1$ is the ESS; e-commerce enterprises will choose to participate in helping farmers.

The dynamic evolutionary process of e-commerce enterprises’ strategic choices is shown in Figure 3. It is easy to see from Figure 2b,c that the increase in government incentives, the reduction in costs, and the increase in benefits from cooperating with consumers to help farmers encourage e-commerce enterprises to participate in poverty alleviation.

4.3. Analysis of the Consumers’ Evolution

The expected payoff of consumers when they choose the “helping” strategy is:

$$\begin{array}{l}{E}_{s1}=f_1\Delta U-P-\Delta P+xyb{f}_2(G+G_3)+x(1-y)b{f}_2(G+G_1)+(1-x)yb{f}_2(G+G_2)\\ +(1-x)(1-y)b{f}_{2}G\end{array}$$

(9)

Similarly, the expected payoff of consumers when they choose the “no helping” strategy is:

$$E_{s1}=-P+xy{f}_2(G+G_3)+x(1-y)f_2(G+G_1)+(1-x)y{f}_2(G+G_2)+(1-x)(1-y)f_{2}G$$

(10)

And the average payoff of consumers is calculated as:

$${\overline{E}}_s=z{E}_{s1}+(1-z)E_{s2}$$

(11)

Hence, the replication dynamic equation of consumers’ choice of strategy is:

$$\begin{array}{l}F(z)=dz/dt=z(E_{s1}-{\overline{E}}_s)\\ =z(1-z)\{f_1\Delta U-\Delta P+(b-1)f_2[y{G}_2+xy(G_3-G_1-G_2)+x{G}_1+G]\}\end{array}$$

(12)

Let $F(z)=0$; there is: $z=0$, $z=1$, $x=x^{*}=\frac{\Delta P-f_1\Delta U-(b-1)f_2(y{G}_2+G)}{(b-1)f_2[y(G_3-G_1-G_2)+G_1]}$.

If $x=x^{*}$, regardless of the value of z, consumers’ strategic choices are always stable. If $x<x^{*}$, we have $\frac{dF(z)}{dz}{|}_{z=0}<0$ and $\frac{dF(z)}{dz}{|}_{z=1}>0$, and thus $z=0$ is the ESS. That is, consumers will choose not to participate in helping farmers and agriculture. If $x>x^{*}$, then $\frac{dF(z)}{dz}{|}_{z=0}>0$ and $\frac{dF(z)}{dz}{|}_{z=1}<0$, implying z = 1 is the ESS. Similarly, consumers will choose to participate in helping farmers and agriculture.

The dynamic evolutionary process of consumers’ strategic choices is shown in Figure 4. The increase in the utility of consumers’ participation in poverty alleviation and the decrease in the premium they pay for additional agricultural products can promote consumers’ active participation in helping farmers and agriculture. In addition, the increase in the quality of agricultural products caused by the government and e-commerce enterprises will affect consumers’ conversion from a negative to a positive attitude towards helping farmers.

4.4. Stability Analysis of Equilibrium Points Based on Replicated Dynamic Systems

Here, we propose the strategic analysis of evolutionary stability under the tripartite interaction. Therefore, a three-dimensional dynamic system is built by copying the dynamic equations of government, e-commerce enterprises, and consumers:

$$\{\begin{array}{l}F(x)=x(1-x)\{z{R}_g+{\alpha }_2{G}_1-(1-\eta )C_g+y[{\alpha }_2(G_3-G_2-G_1)-(1-\eta )T]\}\\ F(y)=y(1-y)\{\eta T+x(T-\eta T)-C_e+[(1-\mu ){\lambda }_1-1]\theta V+z[(1-\mu )\theta ({\lambda }_2-{\lambda }_1)V+R_e-({\lambda }_1-1)\theta V]\}\\ F(z)=z(1-z)\{f_1\Delta U-\Delta P+(b-1)f_2[y{G}_2+xy(G_3-G_1-G_2)+x{G}_1+G]\}\end{array}$$

(13)

Using the simultaneous equation $F(x)=0$, $F(y)=0$, $F(z)=0$, the eight pure strategy local equilibrium points of the system can be obtained as follows: $(0,0,0)$, $(0,0,1)$, $(0,1,0)$, $(1,0,0)$, $(1,1,0)$, $(1,0,1)$, $(0,1,1)$, and $(1,1,1)$. The Jacobian matrix is constructed to judge the asymptotic stability of these local equilibrium points. When the eigenvalue of the Jacobian matrix corresponding to the equilibrium point is negative, the local equilibrium point is the evolutionary stability strategy point. The eight local equilibrium points are brought into the Jacobian matrix, and the eigenvalues are calculated, respectively, as shown in

Table 3. Jacobian eigenvalue of local equilibrium point.

Equilibrium Point	Eigenvalue	Equilibrium Point	Eigenvalue
$(0,0,0)$	${\alpha }_2{G}_1-(1-\eta )C_g$ $(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T-\theta V$ $f_1\Delta U+(b-1)f_{2}G-\Delta P$	$(1,1,0)$	$-[{\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)]$ $-[(1-\mu )\theta {\lambda }_{1}V-C_e+T-\theta V]$ $f_1\Delta U+(b-1)f_2(G+G_3)-\Delta P$
$(0,0,1)$	${\alpha }_2{G}_1-(1-\eta )C_g+R_g$ $(1-\mu )\theta {\lambda }_{2}V-C_e+\eta T+R_e-\theta {\lambda }_{1}V$ $-[f_1\Delta U+(b-1)f_{2}G-\Delta P]$	$(1,0,1)$	$-[{\alpha }_2{G}_1-(1-\eta )C_g+R_g]$ $(1-\mu )\theta {\lambda }_{2}V-C_e+T+R_e-\theta {\lambda }_{1}V$ $-[f_1\Delta U+(b-1)f_2(G+G_1)-\Delta P]$
$(0,1,0)$	${\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)$ $-[(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T-\theta V]$ $f_1\Delta U+(b-1)f_2(G+G_2)-\Delta P$	$(0,1,1)$	$R_g+{\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)$ $-[(1-\mu )\theta {\lambda }_{2}V-C_e+\eta T+R_e-\theta {\lambda }_{1}V]$ $-[f_1\Delta U+(b-1)f_2(G+G_2)-\Delta P]$
$(1,0,0)$	$-[{\alpha }_2{G}_1-(1-\eta )C_g]$ $(1-\mu )\theta {\lambda }_{1}V-C_e+T-\theta V$ $f_1\Delta U+(b-1)f_2(G+G_1)-\Delta P$	$(1,1,1)$	$-[R_g+{\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)]$ $-[(1-\mu )\theta {\lambda }_{2}V-C_e+T+R_e-\theta {\lambda }_{1}V]$ $-[f_1\Delta U+(b-1)f_2(G+G_3)-\Delta P]$

.

First of all, let us assume that any two participants choose not to participate in the process of e-commerce poverty alleviation cooperation; then, the third party will choose not to participate in the poverty alleviation strategy later, that is, the income of participating in e-commerce poverty alleviation cooperation will be less than that of not participating in e-commerce poverty alleviation cooperation. If the conditions of the above assumption are satisfied, which are equivalent to ${\alpha }_2{G}_1-(1-\eta )C_g<0$, $(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T-\theta V<0$ and $f_1\Delta U+(b-1)f_{2}G-\Delta P<0$, it can be seen that the eigenvalues of $(0,0,0)$ are all negative. Therefore, the government, e-commerce enterprises, and consumers do not participate in helping farmers to achieve an ESS. At this time, some eigenvalues of the equilibrium points $(0,0,1)$, $(0,1,0)$, and $(1,0,0)$ are positive, so the situation of these three equilibrium points will not be considered. The stability analysis of the local stability points of the evolutionary game is shown in

Table 4. Evolutionary stability analysis of equilibrium points.

	Case 1		Case 2
Equilibrium Point	Characteristic Root Symbol	Stability	Characteristic Root Symbol	Stability
$(0,0,0)$	$(-,-,-)$	ESS	$(-,-,-)$	ESS
$(1,1,0)$	$(\times ,\times ,+)$	Unstable point	$(\times ,\times ,-)$	ESS when condition (1) is met
$(1,0,1)$	$(\times ,+,\times )$	Unstable point	$(\times ,-,\times )$	ESS when condition (2) is met
$(0,1,1)$	$(+,\times ,\times )$	Unstable point	$(-,\times ,\times )$	ESS when condition (3) is met
$(1,1,1)$	$(-,-,-)$	ESS	$(+,+,+)$	Unstable point

Note: “$\times$” indicates that the current conditions are insufficient to judge the sign.

. Next, we analyze the equilibrium state of the evolutionary game in two cases.

Case 1: When any two of the three parties participate in e-commerce poverty alleviation cooperation, and the cooperative benefit of the third party’s choice to participate in e-commerce poverty alleviation cooperation is greater than the normal benefit of no poverty alleviation, then ${\alpha }_2{G}_1-(1-\eta )C_g<0$, $(1-\mu )\theta {\lambda }_{1}V-C_e+\eta T-\theta V<0$, and $f_1\Delta U+(b-1)f_{2}G-\Delta P<0$ can be obtained. It can be seen that the eigenvalues of equilibrium point $(1,1,1)$ are all negative, and the evolutionary stability strategy points at this time are $(0,0,0)$ and $(1,1,1)$. That is, $(0,0,0)$ and $(1,1,1)$ are ESS.

Case 2: Any one of the three parties achieves more free riding benefits from participating in e-commerce poverty alleviation cooperation through the other two parties. Therefore, when the other two parties choose to actively participate in the agricultural aid, the third party will choose not to participate in the agricultural aid, but will obtain the benefits brought about by the other two parties helping farmers. $R_g+{\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)<0$, $(1-\mu )\theta {\lambda }_{2}V-C_e+T+R_e-\theta {\lambda }_{1}V<0$, and $f_1\Delta U+(b-1)f_2(G+G_3)-\Delta P<0$ can be obtained. Table 3 shows that there are three potential evolutionary stability strategy points, $(1,0,1)$, $(1,1,0)$, and $(0,1,1)$, at this time.

Condition (1): ${\alpha }_2(G_3-G_2)-(1-\eta )(C_g+T)>0$, $(1-\mu )\theta {\lambda }_{1}V-C_e+T-\theta V>0$;

Condition (2): ${\alpha }_2{G}_1-(1-\eta )C_g+R_g>0$, $f_1\Delta U+(b-1)f_2(G+G_1)-\Delta P>0$;

Condition (3): $(1-\mu )\theta {\lambda }_{2}V-C_e+\eta T+R_e-\theta {\lambda }_{1}V>0$, $f_1\Delta U+(b-1)f_2(G+G_2)-\Delta P>0$.

According to the needs of rural revitalization and from the perspective of maximizing the interests of the subjects participating in cooperatively helping farmers, it can be seen that strategy $(1,1,1)$ is a strategic choice that is beneficial to the three parties. The strategies in $(1,1,1)$ are as follows: the government decides to vigorously implement poverty alleviation, e-commerce enterprises choose to participate in poverty alleviation, and consumers would like to help. Because of the needs of the overall development of society, the government needs to focus on agricultural production, promote the stable development of agriculture, and increase farmers’ income. It is important for rural revitalization to promote agricultural products, and to empower and increase efficiency for agriculture with the help of e-commerce enterprises. With the help of e-commerce enterprises and consumers, the government can achieve the goal of helping agriculture; e-commerce enterprises gain benefits and improve their public welfare reputation through government support and consumer help, while consumers can buy the high-quality agricultural products they need, and receive the psychological benefits of helping farmers.

According to the evolution conditions of the stable point, it can be seen that increasing the cost-to-benefit ratio of e-commerce poverty alleviation encourages cooperation among various entities. Setting reasonable rewards for the participation of e-commerce enterprises and having an appropriate agricultural product premium are the key to forming ideal agricultural aid, in the state of $(1,1,1)$.

5. Results and Discussion

In order to analyze the correctness of the above evolutionary equilibrium point and the influence of each parameter change on the strategic choice of each subject, this paper uses Matlab software for simulation analysis, so as to promote the strategic choice of each game participant (vigorously, participation, helping), and form an ideal state of tripartite cooperation for agriculture assistance. According to the conditions of the equilibrium point in Chapter 4.4 and the existing literature [38,40], the parameters from the tripartite evolutionary game model are set to form an array as follows: $R_g=5$, ${\alpha }_2=10$, $C_g=60$, $G_1=4$, $G_2=2$, $G_3=8$, $T=10$, $\eta =0.2$, $C_e=40$, $\mu =0.2$, $\theta =0.2$, ${\lambda }_1=1.2$, ${\lambda }_2=1.8$, $R_e=5$, $V=600$, $f_1=0.6$, $\Delta U=5$, $\Delta P=6$, $b=1.4$, $f_2=1$, $G=4$.

It can be seen that the parameter settings under this array meet the constraints of evolution equilibrium points $(0,0,0)$ and $(1,1,1)$. By substituting the parameters set in the array into the game system to simulate the evolution path of the three parties, with the results of the evolution of the three parties shown in Figure 5, we can see that at this time, the system has two stable points of evolution equilibrium: $(0,0,0)$ and $(1,1,1)$. That is, either the three parties choose to actively participate in helping farmers through cooperation, or none of them participate in helping farmers. Additionally, it can be seen that under the same parameter conditions, different initial probabilities of the three parties will lead to different evolutionary stability results.

In order to intuitively explore the evolution path of the equilibrium point, we promote the development of e-commerce poverty alleviation towards the direction of tripartite cooperation, that is, towards the equilibrium point of $(1,1,1)$. Next, we will discuss the impact of changes in some important factors on the evolution of equilibrium points.

5.1. Influence of the Cost of the Government Choosing the “Vigorous” Strategy and E-Commerce Enterprises Choosing the “Participation” Strategy on Game Evolution

We find that cost is an important factor that affects the decision making of game players [39,48]; therefore, it is necessary to analyze the impact of the e-commerce poverty alleviation cost on the three-way evolution path. We assign the cost of the government’s choice to vigorously implement poverty alleviation as: $C_g=(40,50,71.25,80)$. The cost of e-commerce enterprises participating in poverty alleviation is assigned as: $C_e=(15,30,54.6,60)$.

Figure 6a shows the simulation of the coordinated poverty alleviation strategy when the cost of the government vigorously implementing poverty alleviation takes different values. It can be seen from Figure 6a that when the cost of the government exceeds 71.25, the willingness of the government to help the poor shows a downward trend, driving the decline in the willingness of the entire e-commerce industry to help the poor, finally converging to $(0,0,0)$.

Figure 6b shows the simulation of the coordinated poverty alleviation strategy when the cost of e-commerce enterprises participating in poverty alleviation takes different values. It can be seen from Figure 6b that when the cost of e-commerce enterprises exceeds 54.6, the excessive costs will affect e-commerce enterprises’ poverty alleviation choices, causing e-commerce enterprises to gradually change from participation to no participation. Consumers’ behavioral strategies will also be affected, from “helping” to “no helping”, and the evolution strategy of the three parties will gradually change from $(1,1,1)$ to $(0,0,0)$.

The simulation results in Figure 6 show that the reduction in costs will enhance the willingness of the government and e-commerce enterprises to help the poor. Therefore, scientific and effective cooperative poverty alleviation mechanisms are adopted to control the cost of poverty alleviation within the affordable range of the government and e-commerce enterprises, which may promote the collaborative participation of e-commerce enterprises in poverty alleviation.

5.2. The Influence of the Improvement in Agricultural Product Quality and Agricultural Product Sales Income Caused by Tripartite Cooperation in the Evolution of the Game

In order to analyze the impact of the collaboration benefits generated when the government, e-commerce enterprises, and consumers participate in poverty alleviation on the evolution path of the tripartite system, we assign the improvement coefficients of agricultural product quality and agricultural product income as $G_3=(6,7.1,8,10)$ and $f_1=(1.4,1.65,1.8,1.9)$, respectively. The simulation results are shown in Figure 7.

It can be seen from Figure 7a that with the improvement in the quality of agricultural products brought about by coordinated poverty alleviation from the government and e-commerce enterprises, the evolutionary stability points gradually change from $(0,0,0)$ to $(1,1,1)$. Similarly, in Figure 7b, with the increase in sales revenue brought about by e-commerce enterprises and consumers, the evolutionary stability point also changes to $(1,1,1)$. This is consistent with the fact that the rise in synergy benefits increases the motivation of members to cooperate [36,42].

Figure 7 shows that different cooperation effects will affect the choice of the three parties’ poverty alleviation strategies, and that the improvement in cooperation benefits will promote the collaborative participation of all subjects in e-commerce poverty alleviation. This is due to the synergistic benefits generated by the poverty alleviation actions of all parties being not only beneficial to farmers’ poverty alleviation but also having practical benefits for the participants, which gives them motivation to help the poor. This has promoted the sustainable development of the e-commerce agricultural market.

5.3. The Influence of Consumers’ Preference for Helping Farmers on the Evolutionary Results

Figure 8 provides a simulation diagram of the evolution under the influence of consumer preference for helping farmers. From Figure 8a, it can be seen that the critical value of consumer preference for helping farmers changes between 0.24 and 0.88. When consumers’ preference for helping farmers and agriculture is lower than the critical value 0.24, regardless of the initial probability that the government chooses to vigorously implement poverty alleviation, the tripartite cooperative poverty alleviation tends toward $(0,0,0)$. When the consumer’ preference for helping farmers is higher than the critical value 0.88, no matter whether the government and e-commerce platform choose the strategy of poverty alleviation, consumers will participate in helping farmers. Driven by the enthusiasm of consumers for helping farmers, the probability of government and e-commerce enterprises choosing “vigorously” and “participation” increases, and x, y, and z finally converge to 1.

Previous research emphasized the positive role of consumers’ preference for helping farmers in terms of poverty alleviation, but unfortunately, it only mentioned that consumers’ preference for helping farmers would help improve the synergy between poverty alleviation subjects [19,37], with no further analysis of the evolution of e-commerce poverty alleviation when consumers’ preference for helping farmers was in the middle of its range.

In reality, consumers’ preference for poverty alleviation is generally not at an extremely high or low value. Within the critical value range of (0.24, 0.88), it can be seen from Figure 8b that when consumers’ preference for helping farmers is 0.6, if the probability of the government vigorously implementing poverty alleviation is 0.6, the final evolution result is that the three parties do not participate in helping farmers. Only when the government focuses on its initial strategy of vigorously supporting agriculture and farmers, and guides the social behavior of helping farmers, can it evolve towards the direction of tripartite cooperation in poverty alleviation. When the consumer’ preference for poverty alleviation is 0.8, no matter the initial probability of the government, it will evolve towards the state of tripartite cooperation; however, when the initial probability of the government choosing “vigorously” is low, the probability of the government and e-commerce enterprises choosing to participate in poverty alleviation at an early stage decreases more. Driven by the probability of consumers helping farmers, the willingness of the government and e-commerce enterprises to help farmers gradually increases, and finally evolves to $(1,1,1)$.

Here, we verify that the increase in consumers’ preference for poverty alleviation can promote the poverty alleviation choices of governments and e-commerce enterprises. This is because the helping of farmers by consumers can directly stimulate the agricultural product economy and improve the revenue of e-commerce enterprises and the government. It is further found that when the likelihood of helping farmers is not high among consumers, the government’s guidance on helping farmers at the initial stage is very important. At this time, government intervention can change the evolution result of $(0,0,0)$. When the consumers’ preference for helping farmers is high, the government’s intervention can also accelerate the convergence to $(1,1,1)$.

5.4. The Influence of an Agricultural Products Premium on the Evolutionary Results

Previous research shows that consumers are willing to pay a certain premium for agricultural products [27,47], but there is no quantitative analysis of the relationship between the premium of agricultural products and consumers’ willingness to help the poor. Here, the impact of an agricultural product premium on e-commerce enterprises’ willingness to help the poor is discussed.

Figure 6 provides a simulation diagram of the evolution under the influence of an agricultural product premium. From Figure 9a, it can be seen that the critical value of the agricultural product premium changes between 4.6 and 7.8. When the agricultural product premium is lower than the critical value of 4.6, regardless of the probability of consumers choosing the “help” strategy at the initial stage, the tripartite cooperative poverty alleviation tends towards $(1,1,1)$. When the premium of agricultural products is higher than the critical value of 7.8, the high agricultural premium will encourage e-commerce platforms and the government to choose poverty alleviation strategies in the short term. However, with the decline in the probability of consumers choosing the “no helping” strategy, x, y, and z eventually converge to 0.

Assuming that the premium of agricultural products is within the critical value range, it can be seen from Figure 9b that when the premium of agricultural products is 7, if the consumer’s initial probability of helping farmers and agriculture is 0.6, then it will eventually evolve to $(0,0,0)$. When the probability of consumers initially helping farmers rises to 0.8, this becomes three parties participating in poverty alleviation. Additionally, when the premium of agricultural products is low, the increase in consumers’ initial agricultural aid probability accelerates the convergence to $(1,1,1)$. It can be seen from Figure 9b that when the probability of consumers choosing to help farmers is low at the initial stage, the probability of participation in poverty alleviation by the government and e-commerce enterprises will decline briefly in the early evolution period, and the lower agricultural product premium has a greater impact on the choice of e-commerce enterprises.

Therefore, when e-commerce enterprises find that the probability of consumers choosing to help farmers is low, they can set a lower premium for agricultural products to facilitate tripartite cooperation in helping farmers. When the probability of consumers choosing to help agriculture is sufficiently high, they can obtain a higher income by setting a higher premium for agricultural products.

5.5. The Influence of Incentives for E-Commerce Enterprises to Participate in Poverty Alleviation on the Evolutionary Results

In order to analyze the impact of the government’s incentives for e-commerce enterprises to participate in poverty alleviation, we assign T to (21.25,25) and (10,15). It can be seen from Figure 10a that when T > 21.25, the probability of e-commerce enterprises choosing to participate in poverty alleviation gradually increases in the early stage, but with the decline in the probability of the government vigorously helping farmers, then the tripartite game gradually evolves to $(0,0,0)$, and with the increase in the government’s incentives for e-commerce enterprises to participate in poverty alleviation, the speed of convergence to the three parties’ non-coordination is faster. At this time, the government passively helps farmers, and e-commerce enterprises and consumers do not participate in poverty alleviation activities, leading to the failure of farmers to solve their poverty relief tasks and the failure of rural economy development.

When T < 21.25, the incentives for e-commerce enterprises to choose the “participation” strategy are great enough, and e-commerce enterprises have the motivation to actively participate in helping farmers, which leads to the evolutionary stability point of $(1,1,1)$. It can be seen from Figure 10b that when the incentive for e-commerce enterprises to participate in poverty alleviation is low (T = 10), the state of tripartite cooperation will be reached only when e-commerce enterprises have a high tendency to help farmers (y = 0.8).

Although the government’s incentive mechanism encourages e-commerce enterprises to participate in poverty alleviation [40,41], an excessive incentive is beyond the government’s scope, which is not conducive to the formation of e-commerce poverty alleviation synergy. Figure 10 shows that the government can flexibly set rewards by judging the initial willingness of e-commerce enterprises in advance. When the probability of e-commerce enterprises to help farmers is low, the government can set higher incentives to encourage e-commerce enterprises to adopt agricultural assistance strategies in order to achieve stable tripartite cooperation to help farmers. When the probability of e-commerce enterprises participating in helping farmers is sufficiently high, the government can set a lower incentive value to reduce the financial expenditure of poverty alleviation. Higher agricultural aid rewards can also be set to speed up the evolution to $(1,1,1)$.

6. Conclusions and Suggestions for Future Research

The e-commerce collaborative poverty alleviation model plays an important role in helping farmers out of poverty. However, there are many unstable factors in the coordination mechanism between the government, e-commerce enterprises, and the public. Therefore, this paper regards e-commerce collaborative poverty alleviation as a game system. Based on dynamic game theory, an evolutionary game model is built between the government, e-commerce enterprises, and consumers in the process of e-commerce poverty alleviation cooperation. Through the analysis of the game relationship among the subjects, this paper analyzes the balanced strategy of the three parties, and analyzes the influence of some important factors on the choice of the trilateral cooperative poverty alleviation strategy through simulation. We provide some conclusive support and decision-making suggestions for the sustainable and coordinated development of e-commerce poverty alleviation.

6.1. Conclusions

(1) The reduction in poverty alleviation costs and the increase in the benefits of cooperative poverty alleviation can better encourage the three parties to choose to help farmers. Therefore, by strengthening the interaction mechanism between the government, e-commerce enterprises, and consumers, we can realize the complementary advantages of various poverty alleviation subjects, make rational use of poverty alleviation resources, maximize the benefits of agricultural assistance, and improve efficiency. On the other hand, e-commerce platforms should give full play to the advantages of online digital platforms, catalyze the convergence of the government and consumers to help the poor, and reasonably plan to reduce costs to achieve the goal of poverty alleviation.

(2) Consumers’ preference for helping farmers indirectly affects the strategies of e-commerce enterprises and government for poverty alleviation, and e-commerce enterprises are more sensitive to this factor. The government can act as an advocate for poverty alleviation; when consumers have a low tendency to help farmers, the government should actively assume social responsibility and guide the evolution of the achievements of the trilateral cooperation in poverty alleviation through early, strong support. At the same time, we should encourage and cultivate consumers’ enthusiasm for helping farmers, such as stimulating consumers’ enthusiasm for helping farmers by publicizing the current situation in poor areas.

(3) The government’s incentives and subsidies for e-commerce enterprises can effectively encourage them to participate in poverty alleviation, but excessive incentives increase the government’s financial burden, and only a certain range of incentives can promote the realization of cooperative agricultural assistance. Additionally, the government can set the reward value according to the initial participation probability of e-commerce enterprises. When e-commerce enterprises have a higher tendency to help farmers in the early stage due to their own preference for helping farmers, or in order to obtain a better social public welfare reputation, the government can appropriately reduce the value of incentives; when e-commerce enterprises show a negative tendency towards helping farmers in the early stage, the government can also encourage e-commerce enterprises to participate in helping farmers by setting higher incentives.

(4) Low agricultural product premiums can improve the enthusiasm of consumers to help agriculture. However, due to the e-commerce platform’s pursuit of its own interests and the government’s preference for the rural economy, it is more inclined to set a high premium on agricultural products; thus, the three parties can cooperate to help agriculture only within the appropriate premium range of agricultural products. When consumers have a high probability of helping farmers in the current period, a relatively high premium for agricultural products can be set. If consumers have a low propensity to help farmers in the early stage, they can be guided to participate in poverty alleviation by the setting of a lower premium for agricultural products. Thus, e-commerce enterprises and the government can indirectly increase the price of agricultural products by stimulating consumers’ enthusiasm for helping farmers.

6.2. Practical Suggestions

Collaborative poverty alleviation under the e-commerce model requires the joint participation of the government, e-commerce enterprises, and consumers. Based on the above research conclusions, this paper puts forward some suggestions for promoting the collaborative poverty alleviation mechanism of e-commerce.

(1) The coordination mechanism and the interaction between participants should be improved. First, we should strengthen the effective collaboration between the government and e-commerce enterprises. On the one hand, the government should create a good living environment for the development of enterprises, so that enterprises can both obtain profits and realize social value. On the other hand, e-commerce enterprises can be encouraged to actively participate in poverty alleviation activities by adopting appropriate incentive systems. Secondly, the coordination between the government, e-commerce enterprises, and consumers should be strengthened. The government and e-commerce enterprises can actively stimulate consumers’ enthusiasm for poverty alleviation and use the network advantages of e-commerce enterprises to publicize high-quality agricultural products in poor areas, allowing consumers to understand the importance of the government in poverty alleviation, and stimulating consumers’ awareness of agriculture.

(2) Talent and agricultural product quality in e-commerce poverty alleviation are the key sources of increased income. On the one hand, we should attach importance to the training of talent in the process of e-commerce poverty alleviation, so as to provide positive help and professional guidance for the development of e-commerce poverty alleviation, which greatly reduces the costs and improves the benefits of poverty alleviation. On the other hand, the quality of agricultural products is one of the reasons for consumers’ continuous consumption. We should fully improve the standardization of agricultural products and establish the brand effect of these products in order to further expand the consumption group, which will increase the agricultural economy and achieve the goal of poverty alleviation.

6.3. Future Work

This paper constructs a game model of the government, e-commerce enterprises, and consumers to promote the sustainable development of e-commerce coordinated poverty alleviation. However, in the e-commerce collaborative poverty alleviation model, more subjects can participate, and they will also play an important role in poverty alleviation. In future research, game players such as logistics providers and celebrities on social networks can be introduced to explore the mechanism of coordinated poverty alleviation. At the same time, we hope to obtain actual data using the e-commerce collaborative poverty alleviation model and to restore the real situation to conduct an empirical analysis to obtain more appropriate cooperation suggestions.