Understanding the Role of Rural Poor’s Endogenous Impetus in Poverty Reduction: Evidence from China

Abstract

Motivating the endogenous impetus of the poor to eradicate poverty is an endogenous dilemma that is difficult to solve using the current external poverty alleviation model. In this paper, based on the field survey data of 1112 poor rural households in China, we examine the impact of the poor’s endogenous impetus on their poverty reduction. Firstly, we identify two different components of endogenous impetus: thought impetus and behavior impetus. Secondly, the poverty reduction (livelihood status) of farmers was used as an endogenous variable to construct a partial least squares model to verify our explanation of the role of endogenous impetus of the poor in poverty reduction. The results indicate that (1) both thought impetus and behavior impetus have a positive impact on the livelihood status of the poor; (2) the human capital, physical capital, and social capital of the poor have a positive relationship with the two components of endogenous impetus; and (3) endogenous impetus plays a mediation role between livelihood capital and livelihood status. As expected, human and physical capital have a positive and significant relationship with poverty reduction. The important enlightenment of this study is that it is very important to motivate the poor’s endogenous impetus of escaping poverty in addition to improving external conditions such as livelihood capital owned by farmers in an effort to realize sustainable poverty reduction.

1. Introduction

Eradicating poverty is the primary goal of achieving global sustainable development. In 2000, the Sustainable Livelihoods Approach (SLA) developed by the Department for International Development (DFID) was able to better explain the causes of poverty and its complexity, and is widely used in the research of poverty-related issues [1,2,3,4,5]. The SLA considers that poverty is a livelihood status; and the vulnerability background, the low livelihood capital stock, the lack of policy support, and the weak ability to choose a better livelihood strategy are important factors that turn livelihood status to one of poverty [6,7,8,9]. Among them, livelihood capital (including human capital, physical capital, and social capital) is considered to be a key factor in enhancing the ability of households living in poverty to choose livelihood strategies, increasing their income, improving their living standards, and enabling the poor to escape poverty [4,6,10,11,12,13]. However, most of the existing studies based on the SLA imply the assumptions of “complete information” and “complete rational”, which underestimate the impact of external macro-environmental shocks and the lack of livelihood capital on individual information acquisition and ignore the differences of endogenous cognitive abilities caused by differences in individual livelihood capital stock. Therefore, it is difficult for the poor to make “completely rational” decisions based on the information resources they have. In this regard, the SLA cannot explain the problem of poverty caused by individual subjective factors, namely the endogenous problem of poverty reduction of the poor.

In recent years, with the formation of China’s new era strategic thought on poverty alleviation and development and the orderly advancement of the battle of poverty alleviation, the concept of motivating the endogenous impetus of the poor has been proposed. As stated by Xie and Li [14], endogenous impetus enables individuals to consciously and actively escape poverty. Although there is no consensus on its definition, endogenous impetus is generally explained in the literature as thought impetus and behavior impetus that facilitate the achievement of the goal of sustainable poverty alleviation [15,16,17]. Notwithstanding the difficulties in defining endogenous impetus, existing studies put forward the idea that the endogenous impetus of the poor is an individual’s ability to self-awaken, which may promote the poverty reduction process of the poor [18]. According to Guan et al. [19], the income of poor families is positively related to the cultivation of endogenous impetus. Xue [20] divides the ways of improving the livelihood status of the poor into two categories; one is to enhance the accumulation of livelihood capital and the other is to cultivate endogenous impetus. Liu et al. [21] further point out that, compared with the economic poverty based on the livelihood capital, the subjective poverty of an individual on a psychology level can more effectively reflect the poverty status of the poor. In other words, motivating endogenous impetus is more effective in promoting the poor out of poverty than the accumulation of livelihood capital. The study of Xu [22] shows that the lack of endogenous impetus of the poor has largely slowed down the efficiency of poverty reduction investment and the speed of poverty reduction in China.

Endogenous power is considered as an important factor in the eradication of poverty and a basic guarantee of consolidating the current poverty reduction effect [20]. However, there is no systematic research on how the components of endogenous impetus play a role in the poverty reduction of farmers. Understanding the mechanism by which endogenous impetus affects poverty reduction in the poor will help to reveal how individual endogenous factors contribute to the poverty reduction process of farmers. From this statement, two questions arise: What is the relationship between endogenous impetus and poverty reduction among the poor? And, how does the endogenous impetus promote poor people to escape poverty? The livelihood status of the poor can measure the relative poverty among individuals, that is, whether poverty has been reduced or not. We see this relative reduction in poverty as poverty reduction. Therefore, we focus on the individual internal attributes of the causes of poverty and quantify the endogenous impetus of the poor on the basis of the Sustainable Livelihoods Approach. According to the total sample of 1112 households living in poverty in eight poor counties throughout China, we explore the impact mechanism of endogenous impetus and other types of livelihood capital on poverty reduction of farmers by adopting a structural equation model (SEM), with a view to awaken the individual’s own potential to reduce poverty, breaking the endogenous barriers of poverty reduction and promoting the sustainable poverty alleviation of the poor.

2. Background on Endogenous Impetus and Livelihood Capital

Factors influencing the livelihood status of the poor continue to draw attention from researchers. From the perspective of individual socio-economic characteristics, the accumulation of livelihood capital stock is a key factor for improving the livelihood status of the poor and achieving poverty reduction goals [4,11]. Among them, factors such as the age, gender, education, health protection, labor capacity, and the household size of the farmer represent human capital [23,24,25,26,27]; household assets, productive assets, and infrastructure construction such as transportation, housing, drinking water, energy, and land quality together constitute the physical capital of the farmers [4,23,28]; and social attributes such as social relations, social trust, formal or informal social networks, collective appeals, and opportunities to participate in decision-making are proxies of social capital [10,29,30,31,32]. In addition to these established variables, several authors have emphasized the subjective variables for the main body of poverty reduction, such as mental health, cognitive ability, and risk preference [19,33,34,35], thereby effectively turning the focus to “people” factors (namely endogenous impetus) in addition to external “material” factors (namely livelihood capital) [33,36].

As already mentioned, endogenous impetus has been conceptualized in various ways in the literature. It has been described as the internal tension and driving force of the development of things, and the internal cause of the change of things [14,37]. Since the 1990s, based on the concept of the capacity deprivation of the poor that is emphasized in the Theory of Feasible Ability by Sen [38], the argument of empowerment as a path to reducing poverty in the community has been further proposed. It regards the poor as an individual or family of multiple capabilities and potentials [39]. Lin [40] believes that empowerment as a path to reducing poverty does not empower the poor but is tapping or stimulating the potential of the poor. This potential can awaken the enthusiasm and initiative of the poor in the process of poverty reduction, namely the endogenous impetus of the poor [14,22]. According to Xue [20], the endogenous impetus of the poor refers to the vitality and motivation of the poor themselves, which can effectively promote the improvement of livelihood status. However, from the perspective of psychology and behavioral science, Hang and Hu [15] define “mental poverty” as the result of aspiration failure and behavior failure, which is just the psychological characteristic of the lack of endogenous impetus for the poor [19]. Zhang [17] believes that motivating the endogenous impetus is not only to stimulate the thought impetus to build up the aspiration and confidence of poverty eradication but also to release the behavior impetus of the poor to strengthen their poverty reduction capabilities and rights.

On the spiritual level, evidence shows that the essence of the endogenous impetus of the poor is the exertion of their self-efficacy and the realization of their self-worth [18,41], that is, the poverty reduction aspiration, confidence, courage, and fighting spirit of the poor. On average, psychological variables are often statistically different in people living in poverty compared with people living in affluence [42], which is mainly reflected in the lack of desire and aspiration to change their poverty [43,44]. Aspiration is subordinate to the subjective well-being of the individual in a cognitive dimension and transcends material well-being [45,46,47]. However, the lack of aspiration will inhibit the formation of the individual’s willingness to work [48,49] and will then affect the individual’s economic decision-making and effort level [15,50]. The reduction of the individual’s effort further strengthens the reduction of their aspiration, and the vicious circle formed thereby will cause the individual to fall into the trap of poverty [51]. Xu [37] points out that, relative to the non-poor, the poor show a certain degree of values such as passivity and paralysis, laziness and indifference, and satisfaction with their poverty status, which not only kills the will of the poor to overcome poverty, but also make individuals lose their yearning for a better life. Therefore, motivating the thought impetus of the poor can start with weakening consciousness of poverty, forming an aspiration to escape poverty, building the confidence to have a good life, cultivating the courage and fighting spirit to overcome difficulties, and so on, which promote the poor to change their livelihood status with a positive mental outlook.

The manifestations of the lack of endogenous impetus for the poor are multifaceted. From the perspective of behavioral economics, Hou et al. [52] explain that the main manifestations of the lack of endogenous impetus for the poor are conservative behavior with contradicting poverty alleviation policies, the behavior of being resigned to the present state of assistance after escaping poverty, and the behavior of satisfaction with the poverty status with fear of business risks. The root of these manifestations lies in the long-term scarcity of the poor, leading to a serious imbalance between their intuitive thinking system and their rational thinking system [53]. Individual bounded rationality makes people’s irrational behavior widespread [54], but the loss caused by poor people’s cognitive biases and decision deviations will greatly limit the process of poverty reduction [15]. Banerjee and Duflo [55] find that when the poor who do not have a sufficient food supply receive financial assistance, the criteria for food choice are not price or nutritional value, but how the food tastes. In China, some poor people turn part of the day-old chick that comes from the poverty alleviation project of poultry industry into appetizers the next day, and these chicks should have been used to develop the chicken industry to promote poverty reduction for the poor [19]. Hence, the deviations of cognition and decision-making mean that the poor have lower rational thinking ability than other populations [56], which inhibits the formation of their self-development ability and develops antipathy with poverty alleviation policies [57]. At the same time, as the poor realize that negative shocks can put them into extreme poverty, bankruptcy, or crisis, they tend to avoid risks, resulting in inaction rather than action [58]. Inaction or maintaining the status quo will lead to the loss of opportunities for the poor, which will force them to fall into the trap of vicious poverty for a long time [59,60]. Azam and Gubert [61] take African farmers as an example, proving that relatively wealthy farmers have a stronger spirit of adventure and can obtain a higher income in the future. In view of the many behavioral manifestations of insufficient endogenous impetus, we believe that releasing the behavior impetus of the poor can start with improving the ability of the poor to develop themselves, strengthen their participation in skills training, and increasing their ability to resist risks, which will drive the poor to take the initiative to lift them out of poverty.

According to the literature review, livelihood capital and endogenous impetus are the main factors affecting the livelihood status of the poor. Among them, the endogenous impetus of the poor includes thought impetus and behavior impetus. The absence of aspiration, confidence, courage, and fighting spirit of poverty reduction subjects is the main manifestation of lack of thought impetus. Likewise, the poor’s participation in skills training, self-development ability, and ability to resist risks are the main representatives of behavior impetus. It can be seen that the endogenous impetus of the poor not only pays attention to the awakening of the individual’s spiritual level, but also emphasizes the importance of the individual’s release of behavior impetus. Therefore, this paper defines the endogenous impetus of individuals from the two dimensions of thought impetus and behavior impetus and then explores the poverty reduction mechanism of the endogenous impetus of the poor, and clarifies its role in poverty reduction.

3. Research Hypotheses and Methodology

3.1. Research Hypotheses

Livelihood capital and endogenous impetus together constitute the poverty reduction resources of people living in poverty, which determine their future livelihood results [19], namely, their livelihood status, including whether household income has increased, poverty levels have fallen, and quality of life has improved [10]. However, the poor often have relatively scarce livelihood capital, and their endogenous impetus is severely insufficient. Therefore, there are only two ways to reduce poverty for the poor: One is to promote the improvement of the livelihood status of the poor through the accumulation of livelihood capital; the other is to cultivate endogenous impetus to stimulate the poor to consciously and proactively escape poverty [19]. For this, we proposed the following hypotheses:

Hypothesis 1(H1).

The cultivation of endogenous impetus can effectively promote the improvement of the livelihood status of the poor.

Hypothesis 1(H1a).

Thought impetus has a significant positive impact on the livelihood status of the poor.

Hypothesis 1(H1b).

Behavior impetus has a significant positive impact on the livelihood status of the poor.

As mentioned above, the lack of livelihood capital of the poor, especially the lack of human capital, makes the poor more inclined to limited rationality. The limited rationality then leads to insufficient information and a limited ability to discern information in the decision-making process, which is just the root cause of the endogenous impetus shortage [15,37,52]. Furthermore, the lack of social capital (such as social exclusion and disrespect) will reduce the self-efficacy and aspirations of the poor, and thus inhibit the formation of thought impetus [15,31,34,62]. Evidence shows that the poor can increase their ability to resist risks and individual self-development through the accumulation of livelihood capital, which in turn can stimulate the individual behavior impetus of poverty reduction [17,18,19,20,40]. Based on this, we proposed the following hypotheses:

Hypothesis 2(H2).

The accumulation of livelihood capital can effectively promote the improvement of the endogenous impetus of the poor.

Hypothesis 2(H2a).

Human capital has a significant positive effect on thought impetus and behavior impetus.

Hypothesis 2(H2b).

Physical capital has a significant positive effect on thought impetus and behavior impetus.

Hypothesis 2(H2c).

Social capital has a significant positive effect on thought impetus and behavior impetus.

Generally speaking, the cultivation of endogenous impetus will help farmers fully utilize the advantages of their asset endowment and then make better economic decisions, so as to improve the marginal utility of various types of livelihood capital and improve the livelihood status of the individual [19]. In other words, the cultivation of endogenous impetus can help increase the role of livelihood capital in poverty reduction of the poor. The research of Li, Shuai, Shuai, Cheng, and Liu [10] confirms that the psychological factors of poverty is a mediator variable between social capital and livelihood status, and increasing its level will help strengthen the role of social capital and the improvement of livelihood status and, thus, enable poverty-stricken people to achieve poverty alleviation. Based on this evidence, we proposed the following hypotheses:

Hypothesis 3(H3).

The cultivation of endogenous impetus helps to increase the effect of livelihood capital on the improvement of the livelihood status of the poor.

Hypothesis 3(H3a).

Human capital influences the livelihood status of the poor through the mediating role of thought impetus or behavior impetus.

Hypothesis 3(H3b).

Physical capital influences the livelihood status of the poor through the mediating role of thought impetus or behavior impetus.

Hypothesis 3(H3c).

Social capital influences the livelihood status of the poor through the mediating role of thought impetus or behavior impetus.

In short, the cultivation of endogenous impetus can effectively promote the improvement of the livelihood status of the poor and the accumulation of livelihood capital can effectively promote the increase of the endogenous impetus of the poor; at the same time, the cultivation of endogenous impetus contributes to strengthening the role of livelihood capital in poverty reduction of the poor. Hence, we have constructed a conceptual model of the relationship between livelihood capital, endogenous impetus, and livelihood status of the poor (see Figure 1). Among them, livelihood status indicates the degree of poverty reduction of the poor, which is convenient for understanding the role of the endogenous impetus of the poor in poverty reduction. According to the figure and the hypotheses, livelihood capital is an exogenous variable, while endogenous impetus and livelihood status are endogenous.

3.2. Data and Methodology

3.2.1. Data Description

The data of this research are derived from the field survey questionnaires of eight poverty-stricken counties of China (Changyang County of Hubei Province, Shangcai County of Henan Province, Gonghe County of Qinghai Province, Haiyuan County of Ningxia Hui Autonomous Region, Tongwei County of Gansu Province, Tianzhen County of Shanxi Province, Chahar Right Middle Banner of Inner Mongolia Autonomous Region, and Jinzhai County of Anhui Province) that were carried out by the research group from April to September, 2018. These eight poverty-stricken counties are all nationally designated poor counties, of which Gonghe County of Qinghai Province is a deeply impoverished county (in China, ‘deeply impoverished county’ refers to a county whose poverty rate exceeds 18% according to the Chinese government’s standard for determining the poor population in 2010 (per capita net income of rural family is below 2300 yuan per year)). We used a random sampling method to sample four to five poor villages in each poor county, and then randomly selected about 25 to 35 farmers in each village and used one-to-one field interviews and questionnaires to collect data. The information gathered by the questionnaires mainly included the basic situation of poor families, livelihood capital, endogenous impetus, livelihood status, etc. The questionnaire items mainly used objective numerical and Likert five-point scales. A total of 1255 respondents were interviewed in this field survey, and 1251 effective questionnaires were collected by excluding invalid questionnaires, with an effective rate of 99.68%. As this article focuses on the role of the endogenous impetus of the poor in poverty reduction, the questionnaires for households with a file and card were selected as the research sample, and 1112 questionnaires were obtained. The distribution of the sample area is shown in

Table 1. The details of sample distribution.

Regional Distribution	Changyang County	Shangcai County	Gonghu County	Haiyuan County	Tongwei County	Tianzhen County	Chahar Right Middle Banner	Jinzhai County
Sample Size	123	133	152	113	135	147	144	165

Source: Authors’ calculations according to the questionnaire.

.

The sample shows a high level of diversity among the respondents. The size of the sampled families ranges from one to nine, with an average of three. The family’s per capita net income covers a range of 166.67–65,000 yuan, with an average of 6080.16 yuan. There are still 11.33% families whose per capita net income fall below 2300 yuan. The family daily diet index ranges from 2.1 to 10.5, with an average of 7.06; and the family poverty index differs from 3 to 76, with an average value of 48.352. Among the 1112 registered poor households, 91.01% of the householders are males, and 8.99% are females, covering ages from 18 years to 93 years, with an average of 57. As for educational attainment, 37.6% of the householders have had primary education and 34.1% have had junior and high school education (see

Table 2. The details of variable selection and indicator description.

Latent Variable	Indicators Symbol	Observable Indicators	Indicator Definition	Mean	Standard Deviation
Human Capital (HC)	HC_1	Mean Family Education	Illiteracy-0, Primary education-1, Junior high school education-2, High school diploma-3, University degree-4, Graduate and above-6.	1.324	0.788
	HC_2	Proportion of Peasant-worker	Ratio of peasant-worker to total family size.	0.193	0.225
	HC_3	Proportion of Population Burden	Ratio of the number of individuals without income to total family size.	0.552	0.344
	HC_4	Family Insurance Type	Nothing at all-0, NRCMI and NRSEI-1, NRCMI and NRSEI-3, Commercial insurance-4, NRCMI, NRSEI, and Commercial Insurance-5.	2.927	0.343
Physical Capital (PC)	PC_1	Housing Condition Index	Quantification of two indicators per capita housing area and housing type.	64.134	57.842
	PC_2	Household Fixed Asset Index	Ratio of the amount of assets in the family to total assets of the list.	0.311	0.127
	PC_3	Family Electricity Stability	Very bad-1, Bad-2, Normal-3, Good-4, Very good-5	4.272	0.619
	PC_4	Cultivated Land Quality	Very bad-1, Bad-2, Normal-3, Good-4, Very good-5	3.382	0.855
Social Capital (SC)	SC_1	Leading Role of Village Heads to Escaping Poverty	Very bad-1, Bad-2, Normal-3, Good-4, Very good-5	4.174	0.865
	SC_2	Opportunity to Participate in Democratic Decision-making	Very small-1, Small-2, Normal-3, Large-4, Very large-5	3.897	0.877
	SC_3	Ability to Appeal to Government Agencies for Help	Very small-1, Small-2, Normal-3, Large-4, Very large-5	3.823	0.882
	SC_4	Many Good Friends in the Village	Very small-1, Small-2, Normal-3, Large-4, Very large-5	4.058	0.919
Thought Impetus (TI)	TI_1	Satisfaction with Current Life	Very bad-1, Bad-2, Normal-3, Good-4, Very good-5	4.039	0.856
	TI_2	Confidence in Improving Living Standards in the Future	Very small-1, Small-2, Normal-3, Large-4, Very large-5	4.013	0.863
	TI_3	Frankly Facing Unpleasant Events in Life	Very bad-1, Bad-2, Normal-3, Good-4, Very good-5	4.004	0.856
Behaviour Impetus (BI)	BI_1	Degree of Participation in Skills Training	Very small-1, Small-2, Normal-3, Large-4, Very large-5	2.924	0.922
	BI_2	Self-development Ability	Very small-1, Small-2, Normal-3, Large-4, Very large-5	2.878	1.037
	BI_3	Ability to Resist Risks	Very small-1, Small-2, Normal-3, Large-4, Very large-5	3.574	0.641
Livelihood Status (LS)	LS_1	Family’s Per Capita Net Income	Ratio of total family income (the sum of Planting, breeding, forestry, working, operation, government subsidies and other income) to total family size	6080.16	4506.67
	LS_2	Family Daily Diet Index	The frequency of eating meat, eggs, and vegetables every week, a comprehensive indicator of diet	7.057	1.92
	LS_3	Family Poverty Index	Poverty Scorecard scores	48.352	11.177

Source: a. Proposed by the authors according to the research goal of this study; b. Authors’ calculations; Note: a. NRCMI means New Rural Cooperative Medical Insurance and NRSEI means New Rural Social Endowment Insurance; b. The list of household assets includes agricultural assets, vehicles, household electric appliances, etc., with a total of 12 options. c. Poverty scorecard scores, formulated by the research team based on the Grameen Foundation’s rural poverty scorecard.

).

3.2.2. Variable Selection and Indicator Description

The exogenous variable in the model is livelihood capital, including human capital, physical capital, and social capital. Although the measurement of livelihood capital components is not simple, in the Sustainable Livelihoods Approach human capital is usually measured in terms of education, health security, and labor capacity [24,26] and physical capital is usually measured by factors such as housing, water, energy, land quality, and household assets [1,4,23,28]. Factors such as social trust, collective appeal, and opportunity to participate in decision-making performed well for measuring social capital in a regression model [10,30]. Therefore, human capital is measured by indicators such as mean family education level, labor force share, and family health insurance categories. Physical capital is measured by indicators such as housing condition index, household fixed asset index, electricity stability, and cultivated land quality. Social capital is measured by indicators such as the opportunity to participate in democratic decision-making, the ability to appeal to government agencies for help, and trust among the neighborhood.

As mentioned earlier, the endogenous impetus components include thought impetus and behavior impetus. Thought impetus is measured in terms of “satisfaction with current life”, “confidence in improving living standards in the future”, and “frankly facing unpleasant events in life”. Behavior impetus is measured by indicators of the poor’s participation in skills training, self-development ability, and ability to resist risks. The livelihood status of the poor is measured by indicators of per capita net household income, daily diet index, and poverty index [10]. For details of variable selection and indicator description, see Table 2. Among the latent variables in this paper, the indicators of three latent variables of human capital, physical capital, and behavioral impetus are independent of each other, so they are defined as a formative measurement model, and social capital, thought impetus, and livelihood status are defined as a reflective measurement model.

3.2.3. Research Methodology

For testing the hypotheses, a Structural Equation Model (SEM) was chosen, which allows the identification of the interconnection among endogenous and exogenous variables. We used the livelihood status of the poor (namely poverty reduction) as an endogenous variable to estimate the model. The primary objective of the article is not to understand the factors of poverty reduction per se but rather the role of the poor’s endogenous impetus in the process. For our study, we used a Partial Least Squares (PLS)-SEM model because it performs better in non-parametric analysis [63,64]. Furthermore, as fewer difficulties with multicollinearity occur, it is especially appropriate for studies with intricate causal relationships between the constructs [65]. These estimations were done using Smart PLS Version 3, setting the repeated sampling value to 5000 with Bootstrap technology.

Because all items of our questionnaire were answered by poor households, there may be common method variance in the measurement. This paper uses Harman’s single factor test method, that is, unrotated principal component factor analysis is performed on the terms of all variables in the questionnaire, and the variation of the first factor obtained reflects the amount of common method variance in the data [66]. We used the software of SPSS20.0 to test the sample data. The result shows that the first factor load only accounts for 25.06% in the unrotated principal component factor analysis, which does not account for the majority, so the problem of common method variance in this article is not serious.

4. Results

4.1. The Test Results of the Measurement Models

The test results show that the composite reliability coefficients of the reflective measurement models are above 0.7, and the values of average variance extracted (AVE) are all above 0.5, so the reflective measurement model in this paper has good reliability and convergent validity. For the three reflective measurement models that were elicited, the lowest factor loading among all statements is 0.641, which is above the level considered as the threshold (0.6) [67,68]. The maximum Heterotrait-Monotrait (HTMT) ratio value between latent variables in this paper is 0.648, which is lower than the critical value of 0.85 [69]. Therefore, the reflective measurement models in this paper have good discriminant validity. Among the three formative measurement models constructed in this paper, the outer weights coefficients of each index are significant at the level of 0.01, indicating that each observation variable selected in this paper can effectively represent its latent variable. The variance inflation factor (VIF), which measures the presence of multicollinearity, presents values ranging from one to three, which largely meets the critical value of ten. Consequently, the reflective measurement models and formative measurement models constructed in this paper both passed the reliability and validity tests (see

Table 3. The test results of the measurement models.

Reflective Measurement Models	Reflective Indicators	Factor Loading > 0.7	Indicators Reliability > 0.5	Composite Reliability > 0.7	Average Variance Extracted > 0.5	Variance Inflation Factor
Livelihood status (LS)	LS_1	0.641	0.411	0.806	0.585	1.234
	LS_2	0.723	0.523			1.294
	LS_3	0.907	0.823			1.497
Thought Impetus (TI)	TI_1	0.846	0.716	0.901	0.752	1.781
	TI_2	0.886	0.785			2.097
	TI_3	0.868	0.753			2.019
Social Capital (SC)	SC_1	0.839	0.704	0.867	0.622	1.704
	SC_2	0.850	0.7225			2.006
	SC_3	0.794	0.630			1.797
	SC_4	0.657	0.432			1.329
Formative Measurement Models	Formative Indicators	Outer Weights	T value	P value	Significance Level	Variance Inflation Factor
Human Capital (HC)	HC_1→HC	0.578	13.209	0.000	***	1.101
	HC_2→HC	0.511	11.158	0.000	***	1.235
	HC_3→HC	0.230	4.561	0.000	***	1.156
	HC_4→HC	0.189	4.321	0.000	***	1.002
Physical Capital (PC)	PC_1→PC	0.334	8.164	0.000	***	1.030
	PC_2→PC	0.914	39.617	0.000	***	1.012
	PC_3→PC	0.142	3.139	0.002	***	1.018
	PC_4→PC	0.167	3.886	0.000	***	1.036
Behavior Impetus (BI)	BI_1→BI	0.358	6.537	0.000	***	2.466
	BI_2→BI	0.659	12.506	0.000	***	2.485
	BI_3→BI	0.182	5.201	0.000	***	1.016

Source: Authors’ calculations. Note: a. Bootstrap confidence intervals for 10% probability of error (a = 0.10); b. Significance level: ***: P < 0.01, t-value > 2.57; **: P < 0.05, t-value > 1.96; *: P < 0.1, t-value > 1.65.

).

4.2. The Test Results of the Structural Model

The test of the structure model in this study includes four parts: multicollinearity test, significance test of path, determination coefficient (R² value) test, and predictive relevance (Q² value) test. The results show that there is no multicollinearity in the structural model. Except for the path in which “social capital” has no significant effect on “livelihood status”, the path coefficients between the other latent variables are all significant at the 0.05 level (see

Table 4. The test results of the structural model.

Path	Variance Inflation Factor	Path Coefficients	T value	P value	Significance Level
TI→LS	1.582	0.243	7.019	0.000	***
BI→LS	1.711	0.285	9.932	0.000	***
HC→TI	1.066	0.075	3.198	0.001	***
HC→BI	1.066	0.481	21.158	0.000	***
PC→TI	1.138	0.222	8.216	0.000	***
PC→BI	1.138	0.288	11.030	0.000	***
SC→TI	1.073	0.479	14.543	0.000	***
SC→BI	1.073	0.065	2.472	0.014	**
HC→LS	1.452	0.226	8.818	0.000	***
PC→LS	1.326	0.314	10.375	0.000	***
SC→LS	1.428	0.011	0.398	0.691	NS

Source: Authors’ calculations. Note: a. NS = not significant; b. Bootstrap confidence intervals for 10% probability of error (a = 0.10); c. Significance level: ***: P < 0.01, t-value > 2.57; **: P < 0.05, t-value > 1.96; *: P < 0.1, t-value > 1.65.

). In this model, the R² value of the endogenous latent variable (livelihood status) is 0.596, which indicates that the model has a strong explanatory power and a good fit; and the Q² of the endogenous latent variable is greater than the critical value 0, indicating that the model of the prediction correlation is good (see

Table 5. The test results of determination coefficient and predictive relevance

Endogenous Latent Variable	Thought Impetus (TI)		Behavior Impetus (BI)		Livelihood status (LS)
Determination Coefficient (R2 Value)	0.353		0.402		0.596
Predictive Relevance(Q² Value)	0.256		0.23		0.323
Latent Variable	Effect Size (f²)	Effect Size (q²)	Effect Size (f²)	Effect Size (q²)	Effect Size (f²)	Effect Size (q²)
Human Capital (HC)	0.008	0.004	0.363	0.156	0.087	0.034
Physical Capital (PC)	0.067	0.042	0.122	0.053	0.185	0.059
Social Capital (SC)	0.331	0.210	0.007	0.003	0.000	0.000
Thought Impetus (TI)	-	-	-	-	0.093	0.032
Behavior Impetus (BI)	-	-	-	-	0.117	0.032

Source: Authors’ calculations.

). In summary, both the measurement model and the structural model in our paper passed the test.

The test results of the model (see Figure 2) show that the endogenous impetus components, namely thought impetus and behavior impetus, have a significant positive effect on the livelihood status; the three types of livelihood capital, namely human capital, material capital, and social capital, all have significant positive effect on thought impetus and behavior impetus; and as expected, human and physical capital also have significant positive effect on livelihood status. However, the impact of social capital on the livelihood status is not significant.

4.3. The Results of Mediator Analysis

In order to understand the mechanism of endogenous impetus of the poor on poverty reduction, we further use the Variance Accounted For (VAF) value to test the mediation effect of endogenous impetus components and its type [68].

The test results (see

Table 6. The results of mediator analysis.

Mediator	Path	Path Coefficients	P value	Variance Accounted for Value	Types of Mediator
HC→TI→LS	HC→TI	0.176 ***	0.000	0.144	No mediation
	TI→LS	0.414 ***	0.000
	HC→LS	0.434 ***	0.000
HC→BI→LS	HC→BI	0.568 ***	0.000	0.556	Partial mediation
	BI→LS	0.492 ***	0.000
	HC→LS	0.223 ***	0.000
PC→TI→LS	PC→TI	0.367 ***	0.000	0.205	Partial mediation
	TI→LS	0.323 ***	0.000
	PC→LS	0.461 ***	0.000
PC→BI→LS	PC→BI	0.436 ***	0.000	0.338	Partial mediation
	BI→LS	0.455 ***	0.000
	PC→LS	0.389 ***	0.000
SC→TI→LS	SC→TI	0.543 ***	0.000	0.878	Full mediation
	TI→LS	0.479 ***	0.000
	SC→LS	0.036 NS	0.294
SC→BI→LS	SC→BI	0.185 ***	0.000	0.372	Partial mediation
	BI→LS	0.588 ***	0.000
	SC→LS	0.184 ***	0.000

Source: Authors’ calculations. Note: a. NS = not significant; b. Bootstrap confidence intervals for 10% probability of error (a= 0.10); c. Significance level: ***: P < 0.01, t-value > 2.57; **: P < 0.05, t-value > 1.96; *: P < 0.1, t-value > 1.65.

) show that behavior impetus has a partial mediation effect on the relationship between human capital and livelihood status, namely 55.6% of human capital’s impact on livelihood status is explained via the behavior impetus mediator; but thought impetus has no mediation effect between human capital and livelihood status. As expected, both thought impetus and behavior impetus have a partial mediation effect on the relationship between physical capital and livelihood status. Among them, the mediation effect of thought impetus accounts for 20.5% of the total effect, while the behavior impetus accounts for 33.8%. We also find that thought impetus has a full mediation effect on the relationship between social capital and livelihood status, namely 87.8% of social capital’s impact on livelihood status is explained via the thought impetus mediator. The impact of social capital on the livelihood status is partly realized through the mediation effect of behavioral impetus, which accounts for 37.2% of the total effect.

5. Discussion and Conclusions

The model validates our interpretations of the endogenous impetus components and its role in poverty reduction. We proposed three hypotheses: i.e., the positive association between endogenous impetus and livelihood status (H1), the positive association between livelihood capital and endogenous impetus (H2), and that endogenous impetus plays a mediating role between livelihood capital and livelihood status (H3). In recent years, the importance of endogenous impetus in poverty reduction has often been emphasized in China [40,52]. Our findings on the influence of endogenous impetus on the livelihood status are consistent with many of the results provided in the literature. As Guan, Wang, and Yu [19] found out, the income of poor families is positively related to the cultivation of endogenous impetus. Because we divided the endogenous impetus into two single components, the difference in the degree of impact is recognizable. While both thought impetus and behavior impetus are significant, compared with thought impetus, behavior impetus has a slightly greater impact on the livelihood status. A possible explanation is that an individual’s ability may be more important than their attitude in the process of poverty reduction for the poor. Consequently, participation in skills training, self-development ability, and ability to resist risks improve their livelihoods more, and thus raise them out of poverty. This is in line with the assertion of some authors [17,18].

The livelihood capital of the poor has a significant positive effect on endogenous impetus, probably stemming from the fact that the accumulation of the livelihood capital of the poor may function as a mechanism to decrease or overrule perceptions of risk [19,59], but the different components of the livelihood capital have varying degrees of influence on endogenous impetus. In terms of influence on thought impetus, social capital has the largest impact, physical capital is second, and human capital has the smallest impact. According to the statement used in the questionnaire, social capital refers to the exemplary role of village heads in poverty alleviation, the opportunity to participate in democratic decision-making, the ability to appeal to government agencies for help, and so on. A possible explanation is that the Chinese government’s poverty alleviation effort has been very effective in offering mental education and incentive for the poor. However, compared with physical capital and social capital, human capital has the largest influence coefficient on behavior impetus, a result in line with the literature, which suggest that the lack of endogenous impetus of the poor is mainly due to the shortage of their livelihood capital stock, especially low education in human capital, which has led to the increase in “limited rationality” and “incomplete information” [15,18,37].

Our research results show that endogenous impetus is also a mediator variable for the poverty reduction of the poor, which can effectively promote the impact of livelihood capital on livelihood status. Consistent with Li and Shuai’s [10] findings, thought impetus plays a full mediation role between social capital and livelihood status. The improvement of the level of thought impetus helps to strengthen the role of social capital stock in improving livelihood status, and then promotes the poor to achieve poverty alleviation. The relationships between other types of livelihood capital and livelihood status are also affected to varying degrees by the mediation of different components of endogenous impetus. These results indicate that the cultivation of endogenous impetus is a catalyst for poverty reduction, a view consistent with the literature [19,62]. In addition, as already described in the literature and also substantiated here [26,30,70], in our model, the influence of livelihood capital has to be accepted because both human capital and physical capital have a significant and positive influence on the livelihood status.

This study reveals the compositions of the endogenous impetus of the poor and its role in poverty reduction and explores the poverty reduction mechanism of the endogenous impetus, with the following conclusions.

(1)

Endogenous impetus has a significant positive effect on poverty reduction. Both components of the endogenous impetus have a positive impact on the livelihood status, indicating that the greater the endogenous impetus of the poor, the higher the level of their livelihood, and the better their chance of escaping poverty.

(2)

The livelihood capital of the poor has a significant positive effect on endogenous impetus. Increasing the stock of human capital, physical capital, and social capital of the poor can effectively promote the generation of thought impetus and the release of behavior impetus. In other words, the accumulation of livelihood capital motivates the endogenous impetus of the poor.

(3)

The cultivation of endogenous impetus is a catalyst for poverty reduction of the poor. Endogenous impetus plays a mediation role between livelihood capital and livelihood status, and the increase of its level will help to strengthen the role of livelihood capital and the improvement of livelihood status, and thus enable the poor to get out of poverty.

The implications of this study are multiple perspectives. In China, with the formation of Xi Jinping’s new era strategic thought of poverty alleviation and development and the orderly advancement of the battle of poverty alleviation, poverty alleviation through assisting aspiration and assisting competency is an inevitable requirement for the sustainable development of the poor. Motivating the endogenous impetus of the poor could act as a means to raise people out of poverty with the generation of thought impetus and the release of behavior impetus. Chinese poverty reduction models should combine external poverty reduction with internal poverty reduction. Going in this direction, future poverty alleviation efforts should be built on the programs in which both external poverty reduction strategies are considered, and the cultivation of the endogenous impetus of the poor themselves are paid attention to, in an effort to build a long-term poverty governance model of “two-way hematopoietic” between the poor and the main body of poverty alleviation.

In summary, our study has attempted to understand and estimate the role of endogenous impetus of the poor in poverty reduction. Motivated by the literature on poverty alleviation for the poor, we analyzed the poverty reduction process of the poor embedded in a complex system in which their livelihood capital and endogenous impetus become relevant factors. The interconnectedness of livelihood capital components and the components of endogenous impetus allows us to have a better understanding of how endogenous impetus plays a role in reducing poverty, but there are still open questions. The directionality of the relationships needs more analyses. Furthermore, it is also important to study the construction of endogenous impetus in different contexts and for different poverty reduction methods before extrapolating from these results. What we conclude here is only a first step in quantifying interrelations, if only because the personality and conduct of the poor are likely much more complex than the framework used in this study. Extending that framework to include fragile contexts, shifts in policies and institutions, or other individual characteristics could result in a broader and deeper understanding of the poverty reduction process of the poor.