Analysis of Farm Household Livelihood Sustainability Based on Improved IPAT Equation: A Case Study of 24 Counties in 3 Cities in the Qin-Ba Mountain Region of Southern Shaanxi

Abstract

Sustainable livelihoods are those that are able to cope with and recover from stress and shocks, and that maintain or strengthen livelihood capacity and livelihood capital, without damaging the foundations of the natural environment. In this paper, the IPHACT framework was constructed by improving the classic IPAT equation, and the key factors affecting production output and livelihood sustainability, as well as the factor differences among different livelihood strategy groups, were analyzed. We took 24 counties and districts of Ankang, Shangluo and Hanzhong in the Qin-Ba Mountain area of southern Shaanxi Province as examples, using a survey of farmers’ livelihood status and livelihood capital accounting. The results show that the amplification effect of population size on the environmental impact of livelihood output is widespread and generally significant. Both livelihood sustainability and livelihood benefit passed the significance test in the multi-model analysis, and the negative effect of livelihood sustainability proved the negative correlation between the environmental impact of livelihood output and livelihood sustainability, that is, the higher the livelihood output dependent on natural capital, the greater the environmental impact. The livelihood transformation of Hanzhong City is developing in the direction of reducing the environmental impact of livelihood output, and farmers have successfully practiced green livelihood transformation by changing their livelihood strategies. On the road to common prosperity, livelihood demand will inevitably increase. Reducing the dependence on natural capital is the key to effectively enhancing the sustainability of livelihoods.

1. Introduction

The term “livelihood” has been defined by various scholars, and the Food and Agriculture Organization (FAO), the United Nations Development Programme (UNDP) and the Department for International Development (DFID) have devised a number of frameworks to illustrate its concepts, elements and boundaries. This provides a general definition of “livelihood”, which includes the capabilities, assets and activities needed to earn a living [1]. Livelihood strategy refers to the activities that people choose in order to achieve their livelihood goals according to their human capital, social capital, financial capital, natural capital and physical capital. As the resource base of the family, livelihood capital plays a crucial role in the decision making of the family’s livelihood strategy. From the perspective of the environment, Chambers et al. pointed out that sustainable livelihood is the ability to cope with the impact of pressure and have the resilience to maintain or enhance the original livelihood capacity and livelihood capital without damaging the natural environment on which it is based [2]. Livelihood research has been a key theme in sustainable environmental management [3,4,5,6]. There is a close causal relationship between livelihood and sustainable environmental management, and environmental sustainability cannot be achieved without considering livelihood development.By reducing dependence on the natural environment and eliminating poverty, livelihood activities provide “positive feedback” to the natural environment, while the natural environment maintains human livelihoods and improves human welfare by providing environmental resources and ecological services. Especially for those community residents who rely heavily on the natural environment, livelihood activities demand “negative feedback” from the natural environment. Livelihood, under the concept of sustainable development, cannot endanger the natural environment. Interventions for livelihood improvement and development should focus on reducing the increasing reliance on the environment that is accelerating the destruction of the natural background.

Rural communities dependent on natural resources, especially those dependent on forest resources, such as the direct use of products within the ecosystem environment that consumes natural resources for revenue and ecological compensation, has been the main obstacle to implementing a protection plan [7], as the implementation of a scheme to protect the environment will affect the rural community of peasant household who rely on products from the environment. It has forced local farmers to undergo significant changes in their livelihoods [8]. In addition, farmers need certain factor inputs to maintain their livelihood. Due to the different resources and ecological pressure in their original residential areas, farmers’ dependence on the environment leads to frequent excessive and fast resource consumption behaviors, such as arbitrary land reclamation and deforestation, degradation of the local natural environment, reduction in natural capital and changes of livelihood benefits. This happens more often in developing countries, mainly because the livelihoods of farmers in these regions are highly dependent on income from environmental products. For example, Melaku et al. found that 47% of the total annual per capita income of local people in Ethiopia came from forest products [9]. In Bolivia, 20% of total forest income comes from forest products [10]. This has also been confirmed in the research of other scholars [11,12], including 43.9% in Zambia, 11.59% in Bangladesh, 31.5% in southern China, 15% in Malawi and 24% in Iran [13,14,15,16,17]. In addition, Wang Changhai et al. conducted a survey on farmers near nature reserves in Heilongjiang Province and found that there was a significant difference in the relative environmental income of farmers inside and outside the nature reserves, specifically showing that the income of farmers inside the protected areas were greater than those outside the protected areas [18]. Based on the analysis of the survey data of farmers in seven provinces, including Gansu and Guizhou, Sun Hanlin et al. concluded that natural capital was highly correlated with the livelihood strategies of farmers [19].

It can be seen that income derived from the environment is still very important in the livelihood of residents in developing countries, and farmers are especially dependent on environmental products and services [20,21,22]. To some extent, environmental products play an important role in maintaining sustainable household livelihoods [23]. Agricultural development accounts for 80% of global forest degradation due to environmental degradation and destruction caused by increasing livelihoods. Typical livelihood activities, such as fuelwood gathering, charcoal production, livestock grazing, etc., have accelerated forest degradation. Excessive economic activities of collectors, salt-drying ponds, industrial development and expansion of settlements have become the main factors damaging Asian mangrove ecosystems. The above economic activities intended to meet the needs of human livelihood are increasingly causing degradation and damage to the natural environment. Livelihood strategies pursued by residents/communities, including animal husbandry and forestry, have resulted in overgrazing and overharvesting to varying degrees, and the reality that livelihood activities are unrestrained in their demand for the natural environment still exists. The primary task of intervention, regulation and analysis of human livelihood activities is to reduce the damage of livelihood strategies to the natural environment. In addition, the lack of understanding of local livelihood strategies may lead to the design of inappropriate environmental protection schemes, ultimately leading to unsustainable outcomes such as overuse of resources, livelihood risks and a return to poverty. Sustainable livelihood analysis is both a first step towards reducing environmental pressure and a long-term plan for environmental protection and livelihood enhancement. Agriculture can influence farmers in choosing a livelihood strategy. The output of farmers’ livelihoods will have an impact on natural resources and the agricultural environment. The more dependent on natural resources, the greater the “negative feedback” on the agricultural environment. Sustainable livelihood should not endanger the foundation of the natural environment. The question of how to achieve green livelihoods is closely related to agricultural development, and improving sustainable livelihoods has great significance for agricultural development.

The goal of this paper is to analyze the key factors that affect farmers’ output and livelihood sustainability, as well as the differences in the role of factors among different livelihood strategy groups, and to explore the relationship between livelihood output and livelihood sustainability. On the basis of these, the paper explores the direct or indirect environmental impacts of reducing dependence on natural capital, reducing livelihood activities and strategies, and improving livelihood strategies to effectively enhance sustainable livelihood. In order to illustrate the practical guiding significance of the theoretical analysis framework, this paper builds the IHPACT framework, based on the IPAT classical equation, to analyze and explore the factors that affect livelihood sustainability, as well as the interventions and regulation pathways necessary to achieve livelihood sustainability. The empirical analysis was conducted using the Qin-Ba Mountain area in southern Shaanxi as an example. Based on the above research objectives, Section 2 introduces the main research progress of the IPAT classical formula and expands the analytical framework. Section 3 conducts the result analysis and compares the differences in the roles of different regions and livelihood strategy groups. Section 4 and Section 5 are discussion and conclusion, proposing policy recommendations and future development directions for sustainable livelihoods. Based on the IPAT classical equation, this paper constructs an analytical framework for analyzing and exploring the factors that affect livelihood sustainability, as well as the interventions and regulation pathways needed to achieve livelihood sustainability. The IPHACT framework enriches the application of IPAT classical models in the field of green and sustainable livelihood, not only complementing research on modeling the environmental impact of farmers’ livelihood output, but also providing theoretical and modelling support for environmental protection and long-term improvement of livelihood capacity.

2. Materials and Methods

2.1. Data

In this study, we used participatory rural assessment (PRA) to survey rural households in Ankang, Shangluo and Hanzhong (Figure 1). Then, by calculation, the we analyzed the impact of farmers’ livelihoods on the natural environment. The Qin-Ba Mountain area in southern Shaanxi province includes the cities of Ankang, Hanzhong and Shangluo (

Table 1. Survey area and questionnaire distribution.

Regional	Counties and Districts	Number of Counties/Districts	Effective Questionnaires	Proportion
Ankang	Xunyang, Baihe, Shiquan, Pingli, Ziyang, Langao, Ningshan, Zhenping, Hanyin	9	173	26.9%
Shangluo	Zhen’an, Danfeng, Shangnan, Luonan, Shanyang, Zhashui	6	282	43.9%
Hanzhong	Nanzheng, Chenggu, Yangxian, Mianxian, Xixiang, Lueyang, Zhenba, Ningqiang, Liuba, Foping	9	187	29.2%

). The southern Shaanxi province is located in the hinterland of the Qin-Ba Mountain area, with poor transportation and low economic development levels. This investigation area has large proportion of agricultural output and high dependence on natural resources and the ecological environment. The investigation area covers a larger area, including a variety of livelihoods. At the same time, the composition of local peasant households includes four types, namely, pure peasant households, one-part-households, two-part-households and non-peasant households; farmers’ livelihood strategies can provide strong reference.

From August to September 2020, the research team carried out an actual survey in the Qin-Ba Mountain area of southern Shaanxi province. Using the method of random sampling, 24 sample counties were selected from Ankang, Hanzhong and Shangluo, a total of 796 questionnaires were distributed to farmers, and 642 effective questionnaires were collected, with an effective rate of 80.65%. There were 173 valid questionnaires in 9 counties in Ankang, 187 valid questionnaires in 9 counties in Hanzhong and 282 valid questionnaires in 6 counties in Shangluo. The age distribution of the respondents ranged from 13 to 85. Most of them were illiterate and from rural households with primary education, accounting for 33.68% of the total number of valid applicants. The number of rural households with tertiary education and above was relatively small, accounting for 19.40% of the total number of respondents.

The main content of the questionnaire is the livelihood capital status of the surveyed farmers: it includes five capital types in the framework of the sustainable livelihood analysis, as can be seen from

Table 2. Analysis of economic status of interviewed farm households in the Qin-Ba Mountain area.

Family/Yuan	Pure Households		One-Part Households		Two-Part Households		Non Agricultural
	Number	PER%	Number	PER%	Number	PER%	Number	PER%
Less than 10,000	8	17.78	7	17.07	11	5.42	19	5.38
10,001~20,000	12	26.67	6	14.63	36	17.73	49	13.88
20,001~50,000	10	22.22	13	31.71	82	40.39	96	27.20
50,001~100,000	9	20.00	7	17.07	61	30.05	129	36.54
More than 100,000	6	13.33	8	19.51	13	6.40	60	17.00
Summary	45	100	41	100	203	100	353	100

. Of the different types of farmers that this survey covers, 7.01% of the respondents were “pure households”, 6.39% were “one-part households”, 31.62% were “two-part households”, and 54.98% were “non agricultural”; this shows that there are more people with a high degree of part-time employment in the study area. Among the respondents, 116 people were “college graduates and above”, and 33.58% were “primary school graduates and below”; in addition, 63.40% of the farmers’ annual incomes were between 20,001 and 100,000 yuan, indicating that the income level of most of the sampled farmers was as high as the middle level in the region.

2.2. Methos

The IPAT equation is a conceptual framework equation used to analyze the impact of human activities and the environment through the relationship between the product of three factors, population size, affluence and technology level, and environmental impact. The IPAT equation constructes an identical equation between population and environmental factors [24,25]. As a simple conceptual framework equation, IPAT has the following three characteristics [26,27]: First, it is simple and focuses on humanistic factors, such as population, affluence and technology. Secondly, it systematically analyzes the data relationship between the human drivers of environmental impact and the environmental impact itself. Third, soundness is widely used in the analysis of environmental impacts, such as carbon emissions and water footprint. The development law of natural and social sciences has been obtained by combining natural and social sciences with in-depth exploration on the basis of the IPAT equation. Since IPAT was proposed, it has been widely recognized by researchers in related fields, and has been widely applied in carbon emissions, water footprints and other fields. IGT, ImPACT, STIRPAT and other models have been expanded successively. Based on the classic equation of IPAT, this paper attempts to construct an analytical framework for analyzing and discussing the factors affecting livelihood sustainability, as well as the path for intervention and regulation to achieve livelihood sustainability. Taking the empirical analysis of farmers’ livelihood sustainability in the Qin-Ba Mountain area in southern Shaanxi Province as an example, this paper explains the practical guiding significance of the theoretical analysis framework.

The classic IPAT equation expression is:

$$\mathrm{I}=\mathrm{P}\times \mathrm{A}\times \mathrm{T}$$

(1)

In Formula (1), I is for environmental impact; P is for population; A is for affluence; and T is for skill.

There are some limitations in the analysis of the IPAT equation due to the brevity of the selection of human factors on the right. The variation of factors in the IPAT equation shows a certain proportion, but it does not show a certain proportion in practical application, which limits the application of the IPAT equation in many fields. In recent years, scholars at home and abroad have used IPAT and its extended model to explore the impact of social development on environmental pressure, and it has been gradually recognized and widely used. Scholars such as Harrison have made use of the IPAT equation to assess the environmental impact of fertilizer use, which provides support for studying the impact of technology use on the third agricultural revolution [28]. Mackellar et al. analyzed the validity of the IPAT equation as a tool for assessing energy consumption and CO₂ generation during the period from 1970–1990 [29]. Wernick et al. used the IPAT equation to analyze the number of forests affected by the US wood products industry [30]. Based on the analysis of the relationship between economic growth and the eco-environment in 2010 by Gao et al., the IGT model was constructed [31]. Based on the IPAT equation, Waggoner et al. constructed the ImPACT equation and used it to analyze the impact of the combination of four driving factors on the environment [32]. Schulze closely linked human activity to environmental stress and, on this basis, proposed the construction of the IPBAT equation, in which B represents human behavior [33]. Diesendorf, in contrast, considered that behavioral factors exist in the IPAT equation in terms of population, affluence and technology, so behavioral factors should not be considered as influencing factors in the IPAT equation. According to Xu Zhongmin et al., the evaluation of sustainability using the IPAT equation does not pay attention to the issues of scale, equity and benefit, so the state of social resources (S) is included in their follow-up research, where the IPAT equation is extended to obtain the ImPACTS equation, i.e., I = PACT/S, where S refers to social development and m refers to the impact of management, which itself is one of the components of social resources [34,35]. Using the framework of the IPAT equation, He Qiang and his colleagues analyzed the changes of the ecological environment under the influence of population, economic development, technological progress and economic structure. Zagheni proposed the IPAT equation for a multi-sectoral economy, constructing the population structure according to age, and discussed the insights that might be gained in the context of a stable population theory [36]. Based on the IPAT model, Wang Li et al. analyzed the impact of economic and social development on the environment [37]. Fan Shengyue et al. analyzed the impact of human activities and natural conditions on the environment [38]. Using the STIRPAT model, Li et al. elucidated the influencing factors of carbon emission levels in Tianjin, one of the largest economic centers in northern China [39]. Dietz et al. improved the IPAT equation by adding random variables to it, extending IPAT to the STIRPAT equation [40]; this model has been widely applied to determine the impact of human activities on the ecological environment [41,42,43,44,45], in which STIRPAT is equivalent to IPAT only when a = b = c = d = e = 1. Thus, STIRPA is also known as the IPAT special case. York et al. combined the STIRPAT model with ecological resilience measures to more precisely study the sensitivity of environmental impacts to driving forces and found that the impact of population on carbon emissions was proportional, and that growing economic levels will lead to an increase in carbon dioxide emissions [46]. Based on the basic principle of IPAT, Bai Ling et al. applied the STIRPAT model based on least square regression to study the emission of SO₂, nitrogen oxide and other atmospheric pollutants in our country [47]. Silva et al. put forward the IPAT-e equation and analyzed the impact of listed companies on the environment on the basis of the IPAT-e equation [48]. Yan et al. developed an extensible stochastic environmental impact assessment (STIRPAT) model [49]. Nosheen et al. examined the impact of climate change technologies on green growth in the European economy as a whole and found that in the Eastern and Western European economies panel, over the period 2000–2017, energy-related climate change technologies contributed to green growth, whereas, in the STRIPAT model, environment-related budgets tend to have a positive impact on green growth [50].

Sustainable livelihood analysis can help the main livelihood activities, in full consideration of their own capital occupation, value orientation and the basis of future expectations, to determine the priority of livelihood strategies. Sustainable livelihoods are not a panacea for decision-making and can not be a substitute for decision-support work such as participatory development, sector-wide analysis and integrated analysis of rural development, but sustainable livelihoods do establish a link between livelihood activities and the external environment that influences the outcomes of livelihood strategies. According to the connotation of livelihood and sustainable livelihood, we can find that livelihood activities are individual or family-based (P), and in order to realize the livelihood needs (A,GDP_t), after the balance between the endowment of livelihood capital (L) and the livelihood benefit (T,GDP_t/L), the livelihood output (I) is obtained by determining the livelihood strategy. At the same time, as for the individual, because of the differences of livelihood strategy choice (GDP_a/GDP_t), livelihood capital endowment and livelihood efficiency, the environmental impact of individual livelihood activities (focusing on the natural capital of livelihood capital) is magnified by the number of people (P), such as families, communities and regions, thus affecting local or regional high-quality sustainable development. Based on the IPAT equation, the analytical framework of livelihood sustainability can be obtained as follows:

I (livelihood output environmental impact) = H (livelihood strategy) × P (livelihood subject) × A (livelihood needs) × T (livelihood benefits)

(2)

When I considers to the natural capital of farmers, it can represent the environmental impact of farmers; H reflects the differences in livelihood strategies; C is introduced as an indicator of livelihood sustainability for different livelihood strategy groups, expressed as the ratio of natural capital to livelihood capital; T as livelihood benefit; h as the type of farmer households; P as the number of family members. The above-mentioned framework for the sustainability analysis of livelihoods could be further reformulated to read as follows:

I (natural capital) = h (livelihood strategy group) × P (group population) × A (livelihood needs) × C (natural capital share of livelihood capital) × T (livelihood benefits)

(3)

As the main economic activity subject and basic decision-making unit in rural areas, the input of livelihood capital and the choice of livelihood strategies are closely related to the livelihood benefits of rural households. Farmers’ livelihood benefits are closely related to their livelihood activities, and choice of different livelihood strategies (h); for example, there are differences in the methods and intensities of natural resource use among pure households, one-part households, two-part households and non agricultural natural capital. The irrational livelihood strategies adopted by farmers directly lead to the destruction of the natural environment, which reduces the area of forests and grasslands, and affects the species diversity and ecological restoration function of wetlands; in this way, the large-scale consumption of resources and the change of the natural environment change the impact on the environment, and the livelihood benefit is reduced, which leads to the serious threat to the sustainable development of farmers’ livelihood. The difference analysis of the environmental impact of different livelihood strategy groups can be obtained by the change of farmer household type h, i.e.,:

$$\mathsf{\Delta }I=\left(h_1{}^m-h_2{}^m+\mathsf{\Delta }{h}_n{}^m\right)\times P\times C\times T$$

(4)

Among them, I is the difference of environmental impact under different livelihood strategies, h₁^m and h₂^m are the two livelihood strategies to be compared, and ∆h_n^m is the rebound effect. Among different types of peasant households, such as pure households, one-part households, two-part households and non agricultural, the degree of concurrent employment affects the livelihood benefits and land use level of peasant households, and, in order to promote the positive development of this impact, farmers can adjust the degree of their part-time employment. Agricultural part-time farmers can expand the scope of their non-agricultural activities, enrich the types of non-agricultural activities and choose more appropriate and efficient non-agricultural activities to enhance their livelihood benefits. Non-agricultural farmers can optimize their land use structure, adjusting their planting structure or breeding structure; for example, they can plant mulberry silkworm, walnuts, pepper and so on. At the same time, government departments can gradually establish or improve local non-agricultural industry chains, provide more local employment opportunities for part-time workers and train farmers in cultivation and breeding skills, in order to help them explore more suitable and efficient development paths to enhance the sustainability of their livelihoods. The choice of individual and family livelihood strategy is finally restricted by their livelihood capital and environmental capacity, so that both livelihood activities and livelihood behaviors must deal with the relationship between human beings and social development in order to achieve sustained economic growth and social, resource and environmentally sustainable development. It can be said that the relationship between the features of human livelihood behavior and the resource environment is not only the reflection of the relationship between human and nature, but also causes and promotes the transformation of the relationship between the two. As analyzed above, the change of individual and family livelihood strategies, such as from pure farmers to two-part households, caused the change of livelihood strategies from h₁^m to h₂^m. Furthermore, in the assessment of the change in environmental impacts caused by this change in livelihood strategies (natural resource-based change), it is still necessary to study the ∆h_n^m beside the two livelihood strategies of pure farmers and two-part families in addition to “the difference in livelihood activities”. IHPACT provides the basis for policy adjustment of sustainable livelihoods, and theoretically explains the adjustment of livelihood activities, the change of livelihood behaviors and the leverage of livelihood strategies on livelihood practices. The impact of the government and the market is as follows:

• In terms of government policies, on the one hand, government departments can gradually establish or improve local non-agricultural industrial chains and provide more local employment opportunities for one-part households. On the other hand, the government department can adjust the development strategy, such as transforming idle land with distinctive natural and cultural landscapes into tourist spots according to their characteristics. Replacing the use of natural resources with a new approach makes green livelihoods more coordinated with sustainable development.
• In terms of the market, enterprises can transform resources into assets, funds into equity and farmers into shareholders through agricultural cooperation reform. At the same time, they can also improve the efficiency of natural capital and reconstruct green livelihood capital by developing the understory economy, through methods such as the forest poultry model, forest animal model, forest vegetable model and forest grass model.

However, due to the difficulty in quantifying factors such as government, market and culture, this paper did not select the above factors at present. However, further considering the impact of value orientations (livelihood expectations, social convergence, etc.) and cultural orientations (cultural environment and cultural choice) on livelihood sustainability, this framework can be extended to IHPACTS by further strengthening organizations, institution, etc. Limited by the length of this article, we will not explore this here. In order to further examine the factors affecting the output of farmers’ livelihoods, IPHACT was transformed into a random model according to the formula above, and the specific transformation process referenced [51,52,53]. The transformed stochastic model is treated with a logarithm, and the expression is as follows:

$$\mathit{ln}\left(I\right)=a+b\mathit{ln}\left(H\right)+c\mathit{ln}\left(P\right)+d\mathit{ln}\left(A\right)+c\mathit{ln}\left(C\right)+t\mathit{ln}\left(T\right)$$

(5)

In Formula (5), I represents the environmental impact of the farm household livelihood output; H represents the group characteristics; P is the household population; A represents the farm household livelihood needs; C represents the livelihood sustainability of different livelihood strategy groups. The coefficients of the driving forces (b, c, d, c, t)) represent the percentage of environmental change caused by a change of 1% in the driving forces (H, P, A, C, T). If the coefficient (b, c, d, t) is equal to 1, then the relationship between environmental influence and driving force (H, P, A, C, T) is monotonously proportional, indicating that the acceleration of environmental change caused by increasing this human factor exceeds the change speed of the driving force. If it is greater than 0 and less than 1, this indicates that the increase in the human factor causes environmental change to accelerate at a lower rate than the rate of change of the driving force. If less than 0, it indicates that the increase in the human factor has the effect of mitigating environmental impacts. Secondly, in Formula (5), we defined ln(I) as a dependent variable, ln(H), ln(P), ln(A), ln(C), ln(T), as independent variables, and a as a constant term; regression analysis was carried out on the model after treatment.

From the above analysis, the main factors affecting the farmers livelihood output include the farmers’ livelihood needs, the sustainability of their livelihoods and the characteristics of farmers and technological progress. In general, the main factor affecting the livelihood needs of rural households is per capita household disposable income, and the main factors affecting the characteristics of rural households are the size and type of rural households. The relevant parameters were estimated by using the effective survey data of villages and towns in Ankang, Shangluo and Hanzhong. At the same time, to test whether there is an environmental Kuznets curve hypothesis between farmers’ livelihood needs and environmental impacts, in the following empirical analysis, to avoid the problem of collinearity between A² and A, A² was standardized. The relevant parameter estimates and test results are shown in Table 2.

3. Results

Table 3. Estimated results of environmental impacts of farm households across the region of the survey.

	Constant Term	P	A	A2	C	T	H	w	R2	Sample Size
Model 1	−0.881 *** (−3.16)	0.170 ** (2.32)	0.041 (1.50)						0.1111	642
Model 2	−0.751 ** (−2.49)	0.162 ** (2.20)	0.029 (1.02)	−0.015 (−0.76)					0.1230	642
Model 3	−3.109 *** (−13.67)	1.744 *** (21.09)	1.656 *** (22.90)		−0.397 *** (−12.06)	2.061 *** (24.80)	0.040 * (1.76)	−0.047 ** (−2.13)	0.4147	642

Note: ***, ** and * represent significant at 1%, 5% and 10% statistical levels, respectively. The values in brackets are t values, and w represents regions, the same below.

shows the results of the analysis using different independent variables as models. From Model 1 to Model 3, the estimation results of different influencing factors are given, respectively. The regression result of Model 1 is as shown in the table above. The regression coefficient of population factor is significantly positive at a level of 5%; it can be seen that every 1% increase in population in the study area will increase the environmental impact by 0.17%. The regression coefficient of livelihood demand is positive, but it does not pass the significance test, which shows that the livelihood demands of farmers and the impact of environmental change are moving in the same direction, but the impact is not yet apparent. The magnifying effect of population on the environmental impact of livelihood output is still significant, and because the study area is mostly located in mountain areas, the livelihood and livelihood types of farmers are generally relatively simple, even if more livelihood activities are undertaken to meet higher and higher livelihood needs; relatively stable livelihood strategies do not lead to a significant increase in the environmental impact of livelihood outputs. In Model 2, the quadratic regression coefficient of livelihood demand was added to the basic Model 1. It was obvious that the regression coefficient of population continued to be significant, while the quadratic regression coefficient of livelihood demand of farmers was negative and not significant.

In Model 3, the sustainability of farmers, livelihood benefits, types of farmers and regional differences (w) are introduced. From the regression results of the model, it can be seen that the regression coefficient of population is still significant, the regression coefficient increased from 0.170 in Model 1 to 1.744, which further enhanced the amplification of the environmental impact of population livelihood output; the regression coefficient of livelihood demand was significant at the level of 1%; the regression coefficients of livelihood sustainability, livelihood benefits and regional differences all passed the significance test; and the regression coefficients of livelihood sustainability variables were negative. There is a negative correlation between the environmental impacts of natural capital production and the livelihood benefits, i.e., the higher the dependence on the livelihood output, the greater the environmental impacts, and the significant positive regression coefficients of the livelihood benefits indicate that the higher the livelihood benefits, the higher the environmental impacts caused by the livelihood activities, and that the environmental impacts of their livelihood outputs magnify more than the population size in terms of the regression coefficients; higher value-added livelihood activities bring higher environmental impacts, and the regression coefficient of farmer types is significantly positive, indicating that different types of farmer households have different environmental impacts. Moreover, with the increase in part-time jobs and the complexity of livelihood strategies, the environmental impact will increase, which is consistent with the results of the regression test of the livelihood benefit variables. This shows that the environmental impact of livelihood output under the same conditions has different effects in different regions. In the next part of this paper, we will further analyze and compare the different effects of various factors on different types of farmer households in three cities of southern Shaanxi.

Table 4. Estimation results of environmental impacts of different types of farm households.

Ankang
	Constant Term	P	A	C	T	H	R2	Sample Size
Model 4	−3.194 *** (−9.84)	1.827 *** (14.87)	1.696 *** (15.67)	−0.420 *** (−8.36)	2.120 *** (15.79)	0.181 (0.56)	0.4080	173
Model 5	−2.944 (−3.27)	2.188 *** (4.30)	1.306 ** (2.65)	−0.595 ** (−2.66)	1.979 *** (3.48)		0.5314	10
Model 6	−5.700 *** (−3.30)	1.294 * (1.73)	1.803 *** (3.57)	−0.106 (−0.39)	1.544 (1.81)		0.4356	10
Model 7	−3.023 *** (−4.81)	1.855 *** (7.29)	1.615 *** (6.09)	−0.591 *** (−4.12)	2.222 *** (7.53)		0.3670	48
Model 8	−3.162 *** (−8.74)	1.893 *** (13.86)	1.758 *** (14.31)	−0.415 *** (−7.31)	2.188 *** (14.74)		0.5052	105
Shangluo
	Constant Term	P	A	C	T	H	R2	Sample size
Model 9	−2.588 *** (−6.05)	1.478 *** (10.20)	1.577 *** (10.36)	−0.348 *** (−7.03)	1.974 *** (12.18)	0.082 (1.56)	0.4224	282
Model 10	−2.772 (−1.25)	2.789 ** (2.50)	0.071 (0.06)	−0.674 *** (−3.39)	0.962 (0.80)		0.625	21
Model 11	−2.616 (−0.81)	3.060 *** (7.50)	1.687 *** (6.67)	−0.191 (−1.44)	2.184 *** (4.77)		0.9815	13
Model 12	−2.450 *** (−3.34)	1.315 *** (5.07)	1.408 *** (5.36)	−0.379 *** (−4.19)	1.793 *** (5.77)		0.4562	95
Model 13	−2.144 *** (−4.21)	1.381 *** (6.70)	1.616 *** (7.76)	−0.361 *** (−5.71)	2.037 *** (9.25)		0.4056	153
Hanzhong
	Constant term	P	A	C	T	H	R2	Sample size
Model 14	−0.406 *** (−8.77)	2.116 *** (12.45)	1.761 *** (12.99)	−0.482 *** (−6.40)	2.215 *** (15.40)	0.073 * (1.91)	0.4368	187
Model 15	6.95142	2.694 *** (3.80)	1.771 ** (2.68)	−0.734 *** (−3.66)	2.606 *** (4.40)		0.6996	14
Model 16	−3.573 ** (−1.96)	2.478 *** (4.61)	1.457 *** (3.00)	1.78716	2.370 *** (4.04)		0.5353	18
Model 17	−4.652 *** (−6.37)	2.537 *** (8.04)	1.788 *** (6.79)	−0.630 *** (−3.46)	2.369 *** (7.32)		0.4902	60
Model 18	−3.409 *** (−5.04)	1.761 *** (7.74)	1.749 *** (9.67)	−0.345 *** (−3.75)	2.062 *** (11.54)		0.3804	95

Note: ***, ** and * represent significant at 1%, 5% and 10% statistical levels respectively. The values in brackets are t values, and w represents regions, the same below.

shows the environmental impacts of four different types of farmers’ livelihood outputs in Ankang, Shangluo and Hanzhong. Model 4 represents the environmental impact of all rural households in Ankang on the output of their livelihoods, and Models 5–8 show the relationships between environmental impacts and various factors on the livelihood output of Ankang pure households, one-part households, two-part households and non agricultural, respectively. In Model 4, the regression coefficients of population, livelihood needs and livelihood benefits were positive and significant, and the three factors were still the main magnifying factors of the environmental impact of Ankang livelihood output. However, the coefficient of sustainability variables of farmers’ production was significantly negative, and the regression coefficient of farmer type variables was positive but did not pass the significance test. It can be seen that the general characteristics of the environmental impact of Ankang factors on livelihood output are very similar to that of the sample as a whole. Models 5–8 compared the effects of four different types of farmers in Ankang, and the overall characteristics were similar to that of Model 4: population, livelihood needs and livelihood benefits significantly increased the environmental impact of livelihood outputs, and the sustainability of livelihoods of all types of farmers weakened the environmental impact of livelihood outputs.

Model 9 represents the environmental impact of the total household livelihood output in Shangluo and its relationships with various factors, while Models 10–13 show the relationship between environmental impacts and various factors on the livelihood output of Shangluo pure households, one-part households, two-part households and non agricultural, respectively. In the models, the environmental impacts of population, livelihood needs, livelihood benefits and types of farmers on the livelihood output of farmers in Shangluo are positive, and the impacts of population, livelihood needs and livelihood benefits on the environment are significant: population, livelihood needs and livelihood benefits significantly increased the environmental impact of livelihood output, while the type of farmers had no significant impact on the environment, and the coefficient of sustainability variable of farmers production was significantly negative. The sustainability of the livelihoods of all types of farmers diminished the environmental impact of livelihood output. By comparing Models 9–13, it was found that the comprehensive characteristics of the environmental impacts of various factors in Shangluo on livelihood output are very similar to the overall sample.

Model 14 represents the environmental impact of the total household livelihood output in Hanzhong and its relation to various factors, while Models 15–18 show the relationship between environmental impacts and various factors on the livelihood output of Hanzhong pure households, one-part households, two-part households and non agricultural, respectively. The environmental impacts of population, livelihood needs, livelihood benefits and types of farmers on the livelihood output of farmers in Shangluo were significantly positive, as were the impacts of livelihood benefits to a lesser extent, while the coefficient of the sustainability variable of farmer household production was significantly negative. Similar to the overall characteristics of Model 14, population, livelihood needs, livelihood benefits and household types significantly increased the environmental impact of livelihood output, while the sustainability of the livelihoods of all types of farmers diminished the environmental impact of livelihood output.

As can be seen in Figure 2, the impact on the environment was greatest for pure farmers, who rely mainly on natural capital for their livelihood activities and are the most dependent, while an increase in population would increase the area of arable land used, with the environmental impact of livelihood output consequently increasing. The environmental impact of livelihood needs and sustainability on the output of livelihoods was relatively large, and agricultural production and management are still the main livelihood activities chosen by farmers to meet the increasing livelihood needs. We suggest using natural capital, such as land resources, to promote ecological progress, implementing the “dual carbon” target and implementing a series of national-level strategies such as returning farmland to forests, to some extent. The structure and natural capital of farmers’ livelihood capital have been changed, and the demand for livelihood is bound to increase on the way to common prosperity. Improving the sustainability of the group’s livelihoods is therefore key to reducing the environmental impact of the group’s livelihood outputs. For the two-part households and non agricultural, households, the characteristic was more obvious from the corresponding parameter regression structure. The four types of farmers’ livelihood benefit variables all passed the significance test, and the influence of the variables from pure farmers to non-farmers were increasingly strong and significant. With the gradual reduction in dependence on natural capital and the increasing environmental impact of the indirect production of abundant livelihood strategies, the direct and indirect reliance on and utilization of natural capital for livelihood benefits will be reduced, and this reduction is another key to effectively controlling the sustainability of livelihoods.

Similarly, a comparative analysis in Shangluo and Hanzhong shows that the general trend characteristics of the comparative analysis of different types of farmers in the two regions are very similar: the environmental impact of population on livelihood output is still significantly positive amplification, and increased demand for livelihoods continues to be a dominant factor in the environmental impact of livelihood output, while sustainability of livelihoods is improving steadily, with Hanzhong performing better than Shangluo. In the comparative analysis of different types of farmers in the two cities, the livelihood benefits were significantly positive. In particular, the comparison of the four types of farmers in Hanzhong found that the livelihood benefits showed a downward trend from pure farmers to non-farmers; in contrast to Ankang and Shangluo, the livelihood transition of Hanzhong farmers has reduced the environmental impact of livelihood output, and farmers are successfully implementing a green livelihood transition by changing their livelihood strategies (h). Green livelihood is a kind of sustainable livelihood that does not harm the natural environmental base. Green livelihood capital is a kind of capital that farmers in ecologically fragile areas can depend on. The relationship between green livelihood capital, the ecological environment and natural resources is mutual dependence. Sustainable development must not endanger the natural environment foundation. Interventions for livelihood promotion and development should focus on reducing greater reliance on the environment. By changing livelihood strategies, strengthening environmental awareness and innovating modes of production, farmers can improve natural capital efficiency and rebuild green livelihood capital.

4. Discussion

Research on livelihoods has been an important subject of sustainable environmental management. Increasing research on sustainable livelihoods can improve the ability of farmers, so that they can adjust their livelihood strategy to deal with risks and strengthen their livelihood capacity.

(1)

Through analysis of the survey area, we found that pure farmers and part-time households have a significant impact on the environment. The reasons are as follows: pure farmers and one-part-families are mostly elderly, and their educational level and transportation conditions limit their livelihood choices. Young people choose to leave their homes and work outside, which ultimately lead to the abandonment of arable land and the stagnation of rural development. The government of Hanzhong City has taken great foresight and developed a distinctive tourism industry to provide diverse livelihood choices for young and elderly people. In the future, the government should pay more attention to road construction in remote rural areas and provide skill training services for the remaining population, cultivating a sense of sustainable development, and comprehensively promote the construction of ecological civilization, practice the “dual carbon” goal and achieve sustainable development.

(2)

We selected the ecologically fragile area of Qin-Ba Mountain in southern Shaanxi and found that farmers in the survey area rely heavily on the natural environment. Local farmers consume natural resources too quickly, resulting in a worse fragile ecological environment. This paper analyzes the factors that affect sustainable livelihoods and provides policy recommendations for local farmers to achieve green transformation of their livelihood strategies. Through the study of typical eco-vulnerable areas, it can provide some ideas and relevant reference for areas in developing countries that rely on ecological environment.

(3)

The IPAT model is a classical theory used to study environmental impact, and it has been well applied in carbon emissions, agricultural non-point source pollution, water footprint and other aspects. This article improves the IPAT model and applies it to the theoretical modeling of environmental factors affecting farmers. The framework expands the application scope of the IPAT model and also fills the gap in theoretical modeling research on the environmental impact of farmers’ livelihood output. On this basis, this study can be further expanded to explore the sustainability of rural households’ livelihoods and environmental impact, and also discuss and compare the application of the improved IPAT model in the sustainable livelihood of residents based on registered residence classification, providing theoretical and model support for the long-term improvement of environmental protection and livelihood capacity.

5. Conclusions

In this paper, the IPHACT framework is constructed by improving the classic IPAT equation, and the key factors affecting the sustainability of production and livelihood are analyzed, taking 24 counties in Ankang, Shangluo and Hanzhong as examples. The impact of each factor in the IHPACT framework was analyzed and measured by means of the household livelihood survey and livelihood capital accounting, and the main conclusions are as follows:

(1)

The IPHACT framework model is established to combine the farmers’ livelihood practices with the core factors that affect the environment of livelihood production. This paper probes into the functional relationships between the types of livelihood strategy groups, livelihood needs, livelihood sustainability and livelihood benefits from the theoretical level, and expounds the environmental impacts of the livelihood output of farmers adopting different livelihood strategies. The IPHACT framework provides the basis for policy adjustments in the context of sustainable livelihood development, theoretically explaining the effect of the leverage of livelihood activities, the shift in livelihood behavior, and the improvement of livelihood strategies on the environmental impact of livelihood practices, while further considering the impact of value orientation (livelihood expectation, social convergence, etc.) and cultural orientation (cultural environment and cultural choice) on the sustainability of livelihoods; this analysis framework can be extended to IHPACTS by introducing organizations, institution, etc.

(2)

An analysis of the environmental impacts of four different types of farmers’ livelihoods in Ankang, Shangluo and Hanzhong found that, in Ankang, the impact of the number of pure households on the environment is the greatest, the demand and sustainability of one-part households’ livelihoods have the greatest impact on the environment, and the benefit of two-part households’ livelihoods has the greatest impact on the environment. In Shangluo, the environmental impact of one-part household size, livelihood needs, sustainability and livelihood benefits on livelihood output is the greatest, while the non-farm population has the least impact on the environment of livelihood output. In Hanzhong, the non-agricultural population has the least impact on the environment of livelihood output. Compared with Ankang and Shangluo, Hanzhong has the same impact on the environment of livelihood output as the other three types of farmers. In addition, the quadratic term of farmers’ livelihood needs is inversely related to the environmental impact of their livelihood outputs, but it does not pass the significance test, so it can not be suggested that there is an environmental Kuznets curve between the livelihood needs and the environmental impact on livelihood outputs.

(3)

The impacts of various factors on the environment of livelihood output in Ankang, Shangluo and Hanzhong were investigated. It was found that the environmental impacts of the factors of livelihood sustainability in different livelihood strategy groups were negative; this suggests that the greater the natural capital share in livelihood capital, the greater the environmental impact on livelihood output. In Ankang and Shangluo, the sustainability of pure farmers has the greatest impact on the environmental impact of livelihood output. In Hanzhong, the sustainability factor of one-part household has the greatest impact on the environment of livelihood output. In addition, it also proves that the more dependent farmers are on the environment, the greater the impact on the environment. In particular, a comparison of the four types of rural households in Hanzhong found that the livelihood benefits from pure to non-farm households showed a downward trend, compared to Ankang and Shangluo. Hanzhong farmers’ livelihood transition has reduced the environmental impact of livelihood production, and farmers are successfully implementing a green livelihood transition by changing their livelihood strategies.

Based on the IPAT classical equation, this paper attempts to construct an analytical framework for analyzing and discussing the factors that affect the sustainability of livelihoods and the paths of intervention and regulation needed to achieve the sustainability of livelihoods. Taking the sustainable livelihood of farmers in the Qin-Ba Mountain area in southern Shaanxi as an example, this paper explains the practical significance of the theoretical analysis framework, and finds that:

First, the IPAT model, as a classical theory to study environmental impact, has been applied to carbon emission, agricultural non-point source pollution, water footprint and so on; this paper improves the model and applies it to the theoretical modeling of the environmental impact factors of farmers livelihood output, which extends the application range of the IPAT model. It also makes up the deficiency in domestic research on the modeling of the environmental impact theory of farmers’ livelihood output.

Second, through the above-mentioned analysis, we found that the environmental impact of pure farmers and one-part households fluctuates greatly, which is related to the main income of pure farmers and one-part households dependent on agriculture. In the survey, we found that pure farmers and one-part households are often elderly, living in remote areas far from the city. In this remote rural area, access is limited, and therefore livelihood options are limited, and most local people are less capable, reducing their ability to participate in urban markets. In addition, most of the local farmers have primary school education levels, while some have never been to school and have a low level of education, meaning it is difficult to engage in high-tech jobs. This could be changed by improving rural roads and increasing the level of public transport in rural areas, thereby increasing employment opportunities and diversifying the income of rural households, thereby reducing their dependence on natural capital. In addition, educational development has been strengthened and more high-quality and practical technical education and training courses have been introduced, especially for young people in rural areas, in order to raise their educational and skills levels. To fundamentally raise farmers’ awareness of sustainable development, reduce their dependence on natural capital, reduce the risk of young people choosing a livelihood strategy similar to their ancestors and truly implement the “Dual carbon” target, we will comprehensively promote the building of an ecological civilization.

Third, the environmental impacts in Hanzhong were relatively stable among the three regions studied, which is related to Hanzhong’s pursuit of ecological transformation. For a long time, farmers in Hanzhong have been engaged in inefficient production and inefficient traditional livelihoods, such as rice and traditional Chinese medicine cultivation. In the early days of reform and opening up, policy incentives led to the influx of surplus labor into traditional livelihood activities, and the rise of township and village enterprises increased the standard of living of farmers to some extent, but it also covers up the reality of low production efficiency and poor product benefit. Since the mid-1990s, urbanization has been accelerating, and a large number of young people in rural areas have left their homes and gone to work. Consequently, the phenomenon of cultivated land wastelands and arrested rural development appeared, which was more obvious in the development of mountain areas. Since the 21st century, the Hanzhong government has promoted the diversified development of farmers. As a result, it has created unique tourist attractions such as “10,000 mu of rape flowers” and “1000 mu of tea gardens”, which have boosted the economic development of farmers and promoted local sustainable development. Areas such as Hanzhong can rely on the overall context of ecological construction to develop characteristic agri-tourism in rural areas and encourage more farmers to participate in it, to reduce farmers’ dependence on natural capital and build on existing agricultural development by encouraging the cultivation of distinctive agriculture and the development of green farming, so as to create a diversified tourism landscape. These programs can bring about a reasonable level of change in the type of farm households without destroying the environment or over-consuming resources and promote the achievement of sustainable development goals for farmers.

Later work will strengthen the in-depth research on the sustainability of farmers’ livelihoods and their environmental impacts, and explore the relationship between the sustainability of farmers livelihoods and environmental impacts, as well as sustainable rural development. Based on the IPHACT framework model proposed in this study, the policy regulation mechanism to achieve sustainable livelihoods is explored, and integrated research on the sustainability of farmers livelihoods, environmental impacts and rural sustainable development is carried out, thus providing a scientific basis for the decision-making of promoting rural revitalization and common prosperity.