Content uploaded by Ying Wang
Author content
All content in this area was uploaded by Ying Wang on Jun 05, 2024
Content may be subject to copyright.
J. Geogr. Sci. 2024, 34(6): 1195-1227
DOI: https://doi.org/10.1007/s11442-024-2245-8
© 2024 Science Press Springer-Verlag
Received: 2023-10-11 Accepted: 2024-03-06
Foundation: National Natural Science Foundation of China, No.42371315, No.41901213.
Author: Wang Feifan (1999–), Master Candidate, E-mail: feifanwang5703@cug.edu.cn
*Corresponding author: Wang Ying (1989–), PhD and Associate Professor, specialized in ecosystem services management,
sustainable rural livelihoods, land related policies, and agent-based modeling.
E-mail: yingwang0610@cug.edu.cn
www.geogsci.com www.springer.com/journal/11442
How geographic accessibility and rural governance
mitigate the impact of multiple risks on rural
households’ well-being:
Evidence from the Dabie Mountains in China
WANG Feifan1,2, *WANG Ying1,2
1. Department of Land Resources Management, School of Public Administration, China University of Geosci-
ences, Wuhan 430074, China;
2. Key Laboratory of the Ministry of Natural Resources for Legal Research, Wuhan 430074, China
Abstract: Despite previous studies on the interconnectedness of livelihood risks and hu-
man well-being, limited emphasis has been placed on the influence of geographic acces-
sibility and rural governance on the well-being of rural households. Furthermore, the in-
terplay between geographic accessibility, rural governance, and livelihood risk remains
inadequately explored. Based on 522 household samples collected in the Dabie Moun-
and ordinary least squares regression, this study examines the influence of multiple risks
on the subjective well-being of rural households by investigating the moderating roles
played by geographic accessibility and rural governance. The results show that (1) multi-
risk factors have a significant negative effect on rural households’ well-being (β=−0.219,
p<0.01); (2) geographic accessibility has a weak positive effect on rural households’
well-being (β=0.064, p<0.1) compared to rural governance, which plays a larger positive
role (β=0.228, p<0.01); and (3) geographic accessibility has a significant moderating
on the relationship between multiple risks and rural households’ well-being, decreasing
adverse impact of multiple risks on households’ well-being. Our findings suggest that
geographic accessibility and rural governance have positive implications for enhancing
well-being of rural households. The findings provide policy insight into mitigating livelihood
risks and their negative impacts on household well-being in mountainous regions world-
wide.
Keywords: multiple risks; rural governance; livelihood well-being index; spatial traps; Dabie Mountains, China
1 Introduction
Global poverty reduction is crucial for the well-being of all mankind (Moatsos and Lazopou-
1196 Journal of Geographical Sciences
los, 2021). Due to limited economic development and the relative scarcity of material re-
sources, rural households in developing countries are more susceptible to internal and external
risks (Yasar et al., 2017; Li et al., 2022). As the largest developing country in the world, China
eliminated absolute poverty by 2020 and has entered a critical phase of rural revitalization. It
is crucial to avoid reverting to poverty during this transition period. To effectively link tar-
geted poverty alleviation with rural revitalization, the top priority should be stabilizing pov-
erty alleviation in rural areas and preventing poverty return risks. Because of negative exter-
nalities such as rough terrain, high altitudes, remote locations, poor infrastructure and fragile
ecological conditions (Xu et al., 2020; Wang et al., 2023), mountainous areas are more likely
to experience poverty and weakness (Deng et al., 2023). Rural households in mountainous
areas are often exposed to livelihood risks from various sources that undermine their
well-being and increase their chance of falling back to poverty. Therefore, it is of theoretical
and practical importance to quantify how multiple risks affect rural households’ well-being in
mountainous areas, which could guide targeted policy interventions to promote the sustaina-
ble development of rural society.
Human well-being is a complex concept that has many definitions in different disciplines,
such as wealth, standards of living, and financial security in the field of economics (Wu et al.,
2022); sense of belonging and connectedness in the field of sociology; and health, happiness,
and life satisfaction in the field of psychology (Cai et al., 2023). The Millennium Ecosystem
Assessment (2005) offers a well-known framework for measuring human well-being in five
dimensions: material welfare, health, good social relations, security, and freedom. The Or-
ganization for Economic Co-operation and Development (OECD) measures well-being in two
major domains: material living conditions and quality of life (OECD, 2011). While human
well-being can be measured in different ways, there is a growing consensus that it has both
objective and subjective dimensions (Alatartseva and Barysheva, 2015; Kubiszewski et al.,
2018; Davillas et al., 2022). The objective dimension reflects the material circumstances of
humans (e.g., income, consumption, food, housing) and social attributes (e.g., education,
health, social networks, and connections) (Smith et al., 2019; Zhang et al., 2023b). The sub-
jective dimension captures how humans evaluate their own life satisfaction, happiness, emo-
tions, personal fulfillment, and other psychological states (Liu et al., 2017; Nanor et al., 2021;
Zhang, 2022). However, objective and subjective measures of well-being may not always
align. For example, the well-being paradox suggests that economic growth and income
growth do not necessarily increase individual well-being in the long run. Empirical evidence
also shows that modern material conditions and quality of life have not led to a proportional
increase in people’s well-being (Christoph, 2010). On the other hand, subjective well-being
measures reflect how individuals assess and perceive their quality of life based on their own
criteria, emphasizing their true feelings and subjective heterogeneity. Nevertheless, there is a
scarcity of research examining the subjective well-being of rural households and its spatial
heterogeneity at the microlevel, particularly in environmentally vulnerable and economically
disadvantaged areas. Therefore, in this study, human well-being is measured from a subjective
perspective by evaluating households’ ease of covering living expenses and their satisfaction
with their housing conditions, employment, and life.
Many studies have investigated the impacts of risks on the livelihoods of rural households,
including climate hazards such as temperature extremes, droughts, and floods (Staub and
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1197
Clarkson, 2021; Rajeev and Nagendran, 2023); environmental risks such as wildlife crop
raiding, landslides and soil erosion (Kumar et al., 2020; Zeweld et al., 2020); public health
crises such as the COVID-19 pandemic (Ronkko et al., 2022); and family shocks such as
sudden illness, death, and other unexpected events that can cause labor loss and financial
stress (Arinaminpathy et al., 2009). The current literature on livelihoods and risks focuses on
one type of risk, while relatively few studies have examined multiple risks simultaneously.
Furthermore, although there has been growing interest in mitigating livelihood risks for hu-
man well-being, the majority of studies have focused on livelihood capital (human, physical,
financial, natural, and social capital). Thus, the importance of rural governance and geo-
graphic accessibility is often underestimated.
Rural governance is an interactive process in which multiple actors negotiate and collabo-
rate to achieve public interest when making decisions about public policy or public goods in
rural areas (Sun et al., 2021). Through institutional, economic, and cultural mechanisms, rural
governance can play an effective role in risk aversion, poverty alleviation (Li et al., 2022),
environmental sustainability, social innovation, and community empowerment (Sun et al.,
2021; Palmer et al., 2022; Zhang et al., 2022). During the last decade, rural governance has
received a great deal of attention from both the government and academia (Zhang et al., 2022).
Since 2018, rural governance has appeared in the Central Government’s No. 1 document mul-
tiple times and is considered a powerful way in which to revitalize the countryside (Yin et al.,
2022; Bi and Yang, 2023). Research on rural governance primarily focuses on stakeholders
and actors (e.g., central and local governments, enterprises, social organizations, villagers)
and their roles (Zikargae et al., 2022), governance theory and practice (Zikargae et al., 2022;
Georgios and Barraí, 2023), and effectiveness evaluation (Braga et al., 2023). However, there
is little empirical evidence suggesting that rural governance is correlated with rural house-
holds’ well-being.
Geographic accessibility is a measure of the ability to reach a place (e.g., labor and product
markets, public service facilities, social interaction sites) in relation to another place (e.g.,
rural households’ residence) (Merlin and Jehle, 2023; Mitropoulos et al., 2023; Pot et al.,
2023). This is a reflection of the ease with which destinations can be reached and the oppor-
tunities and inequalities associated with them (Raza et al., 2023), as well as the burden of
traveling between locations (Braga et al., 2023). Geographic accessibility has long been cen-
tral to a host of transportation and regional research and is used as a planning tool (Páez et al.,
2012; Cuervo et al., 2022; Buonocore et al., 2023). Because geographic accessibility deter-
mines the availability of resources and opportunities, it plays a crucial role in human
well-being. The nexus between accessibility and human well-being has been studied in rela-
tion to the role of health care, energy, credit, and the internet in promoting food security, pov-
erty, and subjective well-being (Twumasi et al., 2020; Liang and Li, 2023; Ye and Koch, 2023).
However, there is a lack of comprehensive research on the relationships among rural govern-
ance, geographic accessibility, livelihood risks and human well-being. In particular, less at-
tention has been given to how rural governance and geographic accessibility interact to miti-
gate the adverse impact of multiple risks on rural households’ well-being.
To fill the abovementioned knowledge gaps, this study develops a conceptual framework
that links rural governance, geographic accessibility and multiple risks to human well-being
and quantifies their relationships using 522 household samples collected in the Dabie Moun-
1198 Journal of Geographical Sciences
tains of rural China. Our specific research questions include the following: (1) What is the
impact of multiple risks on the well-being of rural households? (2) How do rural governance,
geographic accessibility and livelihood capital influence the well-being of rural households?
(3) What are the moderating roles of rural governance and geographic accessibility in the re-
lationship between livelihood risks and the well-being of rural households? (4) Does the im-
pact vary among households with different primary sources of livelihood?
This study makes several contributions to the literature. First, multiple risks from both in-
ternal (i.e., health and development risks) and external (i.e., environmental and policy risks)
sources are taken into account. Second, we incorporate geographic accessibility and rural
governance at the community level with livelihood capital at the household level to investi-
gate their impact on rural households’ well-being. Third, both the direct impact of livelihood
risk on rural households’ well-being and its interactive impacts with geographic accessibility
and rural governance are explored. Finally, the heterogeneity in these impacts among house-
holds with different primary sources of livelihood is revealed. The findings of this study pro-
vide policy insight into mitigating livelihood risks and their negative impacts on household
well-being in the Dabie Mountains and other mountainous regions worldwide.
2 Conceptual framework and hypotheses
2.1 Theoretical background
2.1.1 Sustainable livelihoods framework
The sustainable livelihoods framework (SLF), proposed by the Department for International
Development (DFID), has been extensively adopted to better understand the livelihoods of
rural households in developing countries. This framework comprises five interconnected
components: vulnerability context, livelihood capital, transforming structures and processes,
livelihood strategies, and livelihood outcomes (DFID, 1999). The vulnerability context delin-
eates the external environment in which rural households operate, including trends in terms of
socioeconomy, technology, and governance, as well as shocks and seasonality. The frame-
work’s core comprises five types of livelihood capital (i.e., natural, physical, human, financial,
and social) that form the foundation of livelihoods. The transforming structures and processes
within the framework pertain to the institutions, organizations, policies, and legislation that
could influence rural livelihoods. Livelihood strategies refer to the activities and choices that
rural households employ to achieve their livelihood objectives, while livelihood outcomes
represent the achievements or outputs of these strategies. Thus, the SLF provides a framework
for understanding the livelihood processes and outcomes of rural households, which are built
on their capital endowment and influenced by vulnerability contexts and transforming struc-
tures and processes (Guo et al., 2022).
According to the SLF, the vulnerability context constitutes the primary source of risk for
rural households. The accumulation of livelihood capital serves as the foundation for rural
households to make adaptive changes in response to multiple shocks, thereby enhancing their
ability to withstand shocks and improve their livelihoods, ultimately enhancing their
well-being. Livelihood outcomes directly reflect the well-being levels and quality of life of
rural households. Hence, the SLF guides the identification of livelihood risks, the measure-
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1199
ment of well-being indicators, and the selection of factors influencing the well-being of rural
households in this study.
2.1.2 Spatial poverty theory
Spatial poverty theory is a theoretical framework used in social research that measures geo-
graphic spatial disparities (Zhou and Liu, 2022). This theory posits that the geographic distri-
bution of socioeconomic and political factors influences human well-being. Society’s disad-
vantaged groups often cluster in specific geographic areas, constrained by a variety of eco-
nomic, environmental, and social factors. This results in a low level of well-being for resi-
dents of the area, which is a phenomenon that results in what have been termed spatial pov-
erty traps (David et al., 2014). Spatial poverty traps typically occur in geographically remote
and ecologically fragile areas (Liu et al., 2014), where resources and public services, such as
education, medical care, and infrastructure, are inadequately provided. These areas exhibit
high poverty levels, and inhabitants live in relatively poor environments and conditions, lack-
ing opportunities to realize their self-worth, leading to low subjective well-being. Spatial
poverty theory suggests that reducing poverty and inequality in these areas is a vital strategy
for enhancing the well-being and equity of society as a whole.
While poverty alleviation alone may not address all aspects of well-being, it is essential to
recognize that the well-being of rural households is fundamentally intertwined with poverty.
Merely focusing on improving well-being indicators for rural households, without considering
the broader poverty context in which they reside, may result in unstable improvements. Spa-
tial poverty theory allows us to contemplate how differences in geographic factors affect the
well-being levels of residents in spatially deprived regions. It also poses the following intri-
guing question: does geographic accessibility play a role in mitigating the impact of multiple
risks on rural households? Thus, this study investigates the direct effects of geographic acces-
sibility on the well-being of rural households and the moderating effects of geographic acces-
sibility on the livelihood risks and stresses faced by rural households, drawing on spatial pov-
erty theory.
2.1.3 Rural governance theory
Rural governance theory provides a framework for understanding and improving public poli-
cies and institutions that influence rural development. The theory emphasizes the interaction
and cooperation among various actors and levels of governance, such as local communities,
civil society, the private sector, and public authorities (Esparcia and Abbasi, 2020; Georgios
and Barraí, 2023). It also underscores the territorial approach, acknowledging the diversity
and uniqueness of rural areas and their developmental potential. Rural governance theory
aims to address challenges such as natural resource management, rural marginalization, un-
sustainable transformational rural development, and ecological degradation of the countryside.
To foster rural development, innovative rural governance approaches, including decentralized,
participatory, social network, and distributed governance approaches, have been introduced,
all of which have significantly contributed to rural revitalization (Zhang et al., 2022; Dansou
and Carrier, 2023). Previous comparative studies of rural areas in the European Union (EU)
have highlighted the importance of social governance in stimulating the endogenous dynam-
ics of rural development and revitalizing rural areas (Georgios et al., 2021).
1200 Journal of Geographical Sciences
The rural governance structure in China is a practical form of rural social relations and so-
cial management (Ye and Liu, 2020). It requires constant innovation to adapt to the rapidly
evolving rural economy. After the reform and opening up, China embarked on a new era of
socialist modernization. State power receded from rural areas, leading to decentralized rural
governance. Township governments and village committees became the paradigm of rural
politics, supervising and guiding grassroots self-governance organizations, such as villagers’
committees, in managing village public affairs. Currently, China’s rural areas have a grass-
roots social governance framework led by the Party, dominated by villagers’ autonomous or-
ganizations and people. Rural governance is transitioning toward shared governance involving
multiple actors. This governance model respects and safeguards the status of rural households
as the primary entity, stimulating households’ vitality by leveraging local individuals to guide,
inspire, and activate various public resources in rural areas (Min, 2020). This model also cap-
italizes on the strengths and efforts of multiple actors to unearth resources for rural revitaliza-
tion. Drawing upon rural governance theory, the governance capacity of rural villages has
emerged as a significant determinant of rural households’ well-being; thus, it is integrated into
the conceptual framework of this study.
2.2 Measures of rural households’ well-being
In this study, we select the following four indicators to measure rural households’ well-being
from a subjective perspective: ease of covering living expenses, satisfaction with housing
conditions (Kearns and Whitley, 2020), job satisfaction (Amfo et al., 2023) and life satisfac-
tion (Easterlin et al., 2011). We then compute a composite satisfaction score, i.e., the Liveli-
hood Well-being Index (LWI), by assigning weights to each indicator to characterize the sub-
jective well-being of rural households (Table 1). To determine the weight for each indicator,
we integrate the entropy weight method (EWM) and the analytic hierarchy process (AHP).
Table 1 An index system for the outcome variable: rural household well-being
Variable name Description Mean SD Weight
Ease of covering living expenses Score of ease of covering living expenses
(1=very difficult to 5=very easy) 3.141 0.909 0.364
Housing condition satisfaction Score of satisfaction with housing conditions
(1=very low to 5=very high) 3.729 0.890 0.134
Job satisfaction Score of job satisfaction
(1=very low to 5=very high) 3.457 0.869 0.093
Life satisfaction Score of life satisfaction
(1=very low to 5=very high) 3.617 0.889 0.408
The ease of covering living expenses serves as an indicator of a household’s economic sit-
uation. In general, the easier it is to cover living expenses, the better the household’s econom-
ic situation is. The level of satisfaction with housing conditions reflects the material condi-
tions of the household. Higher job satisfaction levels are associated with greater social partic-
ipation and social identity among family members. This can reduce their psychological stress
and sense of social injustice, which in turn can reduce the negative impact on their health sta-
tus and lead to a greater level of life satisfaction. Therefore, households that can easily cover
living expenses and express satisfaction with their housing, job, and life conditions generally
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1201
exhibit a greater level of subjective well-being. Thus, these four indicators are selected to
measure the subjective well-being of rural households.
2.3 Influencing factors of well-being and hypotheses
2.3.1 Multiple risks
Livelihood risk is a key variable that negatively affects rural households’ LWI. This study
considers multiple risks from both internal (i.e., health and development risks) and external
(i.e., environmental and policy risks) sources (Table 2). Health risk refers to factors that
threaten the health status and security of family members or livestock and is represented by
three indicators, i.e., family member illness, lack of medical insurance (Zeng et al., 2021) and
accidental livestock loss (Wang et al., 2020a). Developmental risks are uncertain events or
circumstances that can adversely affect households’ pursuit of sustainable livelihoods and are
manifested by the following four indicators: high expenses in house building and renovation,
education, social gifts, and medical expenses. Environmental risk refers to natural disasters or
shocks that can directly affect agricultural production and threaten food security (Thapa, 2010)
and is measured by the following three indicators: natural disasters (Kuang et al., 2019), wild-
life crop raiding (Thapa, 2010), and crop diseases and insect pests (Kuang et al., 2019). Policy
risk refers to the unintended disturbance and uncertainty of policy interventions that may af-
fect rural livelihoods (Wang et al., 2021b) and is measured by participation in payment for
ecosystem services (PES) programs (Xiao et al., 2023), high dependence on transfer payments,
and lack of social security.
Based on the SLF, multiple risks from both external and internal sources shape the vulner-
ability context and influence the accumulation of livelihood capital, all of which contribute to
negative livelihood outcomes. Existing studies in the literature have shown that rural house-
holds, when exposed to shocks and risks, tend to deplete their financial and durable assets
(Liu et al., 2023), reduce their consumption (Qiu et al., 2023), decrease their agricultural in-
vestment (D’Exelle et al., 2024), and adopt other strategies to smooth out these effects (Ghaz-
ali et al., 2022; Ghazali et al., 2023). These actions may further lead to malnutrition (Jan-
kowska et al., 2012), health issues (Roy et al., 2024), a return to poverty (Alemie et al., 2022),
reduced productivity (Tran et al., 2022), and increased vulnerability (Kumar et al., 2020). All
these factors contribute to diminished well-being. Thus, based on the SLF and the literature,
we propose the following hypothesis:
Hypothesis H1: Multiple risks have a significant negative effect on the well-being of rural
households.
2.3.2 Geographic accessibility
Geographic accessibility can be measured based on travel time, travel cost, and spatial dis-
tance (Páez et al., 2012). In this study, travel time, which is the reported commute time from a
rural household’s residence to a specific public service location by the most frequently used
means of transport, is used to estimate geographic accessibility. Travel time can provide more
information than distance, as it reflects local geographical conditions, rural households’ per-
sonal travel preferences, and their physical capital (i.e., transportation tools). As shown in
Table 2, the travel times to the nearest county center, township center, primary school, sec-
ondary school, and medical site are selected to measure geographic accessibility. These indi-
cators capture rural households’ access to labor and product markets and major public service
1202 Journal of Geographical Sciences
facilities (Wang et al., 2020b).
According to spatial poverty theory, poverty is significantly determined by disadvantaged
locations, such as mountainous areas characterized by low accessibility to paved roads, major
infrastructure, and public services (van Dülmen et al., 2022). In contrast, households located
in advantageous geographical locations, such as those near main roads and health care facili-
ties, can receive more immediate assistance when facing shocks and risks, thereby mitigating
their negative impact on well-being (Zhang et al., 2023a). Therefore, the following hypotheses
are proposed:
Hypothesis H2: Geographic accessibility has a positive impact on rural households’ LWI.
Hypothesis H3: Geographic accessibility plays a negative moderating role in the relation-
ship between multiple risks and rural households’ LWI, decreasing the negative impacts of
multiple risks.
2.3.3 Rural governance
Rural governance can be assessed from the perspective of the living environment, social eq-
uity, social security, public services, fairness in resource allocation, and social participation
(Halla et al., 2022; Shen et al., 2023). In this study, seven indicators are selected as measures
of rural governance, namely, the operation of rules and regulations, road cleanliness, village
greenness, cultural and entertainment service satisfaction, employment service satisfaction,
fairness in resource allocation, and access to collective information (Table 2). The operation
of rules and regulations can reflect the degree of village organization and coordination. Road
cleanliness and village greenness manifest the effectiveness of village environmental man-
agement. Satisfaction with cultural and entertainment services, and employment services re-
flects the level of village public service provision. In addition, fairness in resource allocation
and access to collective information are important indicators that characterize the degree of
civilization in society, which is also an important aspect of rural governance.
Good governance practices, including transparent and orderly regulation operations, robust
service provision, effective environmental governance, and extensive access to employment
information, can help meet the needs and rights of rural individuals and enhance their
well-being. Specifically, effective regulation operations can ensure fair resource allocation,
while robust service provision can maximize benefits to the general population, improve ac-
cess to basic services such as health care, education, and clean water, and enhance rural
households’ satisfaction with their livelihoods, employment, health, and other aspects of their
overall well-being in multiple dimensions. In addition, villages that demonstrate robust rural
governance capacity can actively assist rural households in building resilience against risks
and provide immediate aid to facilitate recovery from unexpected shocks (Wang et al., 2024).
This can be achieved through the establishment of mutual-aid social networks and the creation
of community-based emergency funds or insurance schemes (Wang et al., 2021a).
Therefore, based on rural governance theory and the above analysis, we propose the fol-
lowing hypotheses:
Hypothesis H4: Rural governance has a positive impact on rural households’ LWI.
Hypothesis H5: Rural governance plays a negative moderating role in the relationship be-
tween multiple risks and rural households’ LWI, decreasing the negative impacts of multiple
risks.
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1203
Table 2 Explanatory variables: livelihood risks, geographic accessibility, and rural governance
Va ri a b l e
category Variable name Description Mean SD Weight
Environmen-
tal risk
Natural disasters
Severity of natural disasters such as floods, droughts,
debris flows, soil erosion, and abnormal weather
(0=very less serious to 5=very serious)
0.726 1.540 0.527
Wildlife crop raiding Severity of crop damage by wildlife
(0=very less serious to 5=very serious) 0.910 1.639 0.136
Crop diseases and
insect pests
Severity of crop pests and diseases
(0=very less serious to 5=very serious) 0.410 1.031 0.337
Policy risk
Participation in
PES programs
Whether participates in PES programs that require
rural households to give up their right to use
natural resources (1=yes, 0=no) 0.437 0.496 0.070
High dependence
on transfer payment
Percentage of government transfer payments in
household income (%) 1.395 0.905 0.258
Lack of social security Whether all family members have not included
in the social security system (1=yes, 0=no) 0.090 0.287 0.672
Health risk
Family member
illness Number of sick household members 0.163 0.445 0.562
Livestock
accidental loss
Whether experienced accidental loss of livestock
(1=yes, 0=no) 0.054 0.226 0.436
Lack of medical
insurance
Percentage of household members not covered by
health insurance 0.966 0.154 0.002
Development
risk
High expense in house
building and renovation
Amount of building and renovation expenses in
the past 12 months (yuan) 4,046 22,656 0.614
High expense in
education
Expenditures on education in the past 12 months
(yuan) 5,917 9,476 0.364
High expense in
social gifts
Expenditures on social gifts in the past 12 months
(yuan) 6,940 6,640 0.004
High
medical expenses
Amount of medical expenses in the past 12 months
(yuan) 5,319 11,882 0.017
Geographic
accessibility
Travel time to
county center
Travel time to the nearest county center by
transport means usually used (minutes) 45.510 26.970 0.485
Travel time to
township center
Travel time to the nearest township center by
transport means usually used (minutes) 19.700 11.850 0.126
Travel time to
primary school
Travel time to the nearest primary school by
transport means usually used (minutes) 17.470 10.790 0.035
Travel time to
secondary school
Travel time to the nearest secondary school by
transport means usually used (minutes) 19.230 11.260 0.066
Travel time to
medical site
Travel time to the nearest medical site by
transportation means usually used (minutes) 13.300 9.574 0.287
Rural
governance
Operation of rules
and regulations
Score of operation of rules and regulations
(1=very poor to 5=very well) 3.497 0.735 0.238
Road cleanliness Score of road cleanliness (1=very poor to 5=very well) 3.603 1.084 0.050
Village greenness Score of greening degree of the village
(1=very poor to 5=very well) 3.873 1.042 0.024
Cultural and
entertainment services
satisfaction
Score of satisfaction with cultural and
entertainment services (1=very poor to 5=very well) 3.042 1.043 0.165
Employment
service satisfaction
Score of satisfaction of labor employment service
(1=very poor to 5=very well) 3.017 0.951 0.094
Fairness in
resource allocation
Score of fairness in resource allocation
(1=very poor to 5=very well) 2.996 1.080 0.411
Access to
collective information
Score of the access to information about collective
affairs (1=very poor to 5=very well) 3.811 1.027 0.016
1204 Journal of Geographical Sciences
Table 3 Control variable: livelihood capital
Category Variable name Description Mean SD
Human
capital
Infants Whether has infants and toddlers aged under three years old
(1=yes, 0=no) 0.048 0.214
Preschool children Number of preschool children aged between three and seven years old 0.156 0.414
Students Number of household members who are receiving different levels
of education from elementary school to graduate school 0.590 0.815
Elderly members Number of seniors aged above 65 years old 0.731 0.825
Labor force Number of household members who can contribute to production
or income growth 2.218 1.163
Highest education
Highest education level among family members (1=never attended
primary school; 2=primary school; 3=middle school; 4=high school/
vocational high school/secondary school; 5=postsecondaryschool/
bachelor’s degree; 6=postgraduate degree)
3.483 1.214
Mean health Average health status of household members (1=very unhealthy
to 5=very healthy) 4.197 0.850
Natural
capital
Paddy area The area of the contracted paddy field (mu) 2.315 1.409
Dryland area The area of contracted dryland (mu) 1.293 0.999
Water quantity Whether rivers and reservoirs can meet irrigation needs (1=yes, 0=no) 3.552 1.067
Water quality Score of quality of water sources (1=very poor to 5=very good) 3.707 1.013
Physical
capital
House type
Score of house type (1=earthen-walled and tile-roofed house;
2=single-story brick house
3=two-story house without indoor bathroom; 4=two-story house with
indoor bathroom
5=three-story house and above)
4.263 0.768
Energy use
Score of fuel type (1=only firewood; 2=mainly firewood, occasional-
ly coal gas/liquefied gas/natural gas; 3=half of coal gas/liquefied
gas/natural gas, half of firewood; 4=mainly coal gas/liquefied gas/
natural gas, occasionally firewood; 5=only coal gas/liquefied gas/
natural gas, no firewood)
3.695 1.088
Sanitation Score of sanitary facilities (1=outdoor public toilet; 2=outdoor dry
toilet; 3=outdoor flush toilet; 4=indoor flush toilet) 3.637 0.869
Electrical appliances
Weighted number of electrical appliances (the weights of air
conditioner, water heater, refrigerator, washing machine, rice
cooker are 0.3, 0.3, 0.2, 0.1, 0.1, respectively)
1.083 0.560
Entertainment
equipment
Weighted number of communication and entertainment devices
(the weights of computer/pad, cell phone, TV, and stereo/radio
are 0.4, 0.2, 0.25, 0.15, respectively)
1.037 0.469
Farm tools
Weighted quantity of farm tools (the weights of tractors/transport
tractors, threshing machines/other small handlers, electric water
pumps, oxen/mules, hoes, other small farm implements are 0.4, 0.25,
0.1, 0.05, 0.15, respectively)
0.354 0.318
Transportation tools
Weighted number of vehicles (the weights of cars/minivans, small
trucks, motorcycles/motorcycles, battery cars, bicycles/pedal tricycles
are 0.35, 0.25, 0.2, 0.15, 0.05, respectively)
0.557 0.251
Financial
capital
Savings Whether has savings (1=yes, 0=no) 0.503 0.499
Insurance
Whether purchases commercial insurance, such as commercial medi-
cal insurance, motor insurance, property insurance, commercial life
insurance, etc. (1=yes, 0=no)
0.305 0.461
Financial products Whether has financial products, such as stocks, funds, government
bonds, trust products, foreign exchange products, etc. (1=yes, 0=no) 0.012 0.107
Income
Annual household income range (1=under 10,000 yuan; 2=10,000–
50,000 yuan; 3=50,000–100,000 yuan; 4=100,000–150,000 yuan;
5=150,000–200,000 yuan; 6=over 200,000 yuan)
2.762 1.179
(To be continued on the next page)
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1205
(Continued)
Category Variable name Description Mean SD
Social
capital
Work in village
collective Whether family members work in the village collective (1=yes, 0=no) 0.073 0.260
Wo rk i n lo ca l
government
Whether family members, relatives, or friends work in the county
or township government (1=yes, 0=no) 0.146 0.353
Participation
in collective affairs
Score of participation in collective affairs
(1=very low to 5=very high) 2.578 1.209
Relations with
villagers
Score of relations with villagers
(1=very estranged to 5=very close) 3.864 0.740
Ease of borrowing How easy it is to borrow from others when in need of a large
amount of money (1=very difficult to 5=very easy) 3.283 1.045
Number of helpers Number of people who offer help when household encounters
difficulties (persons) 4.006 0.844
Notes: 1 US $=7.14 yuan in 2023.
2.3.4 Control variables: livelihood capital
Livelihood capital is an important dimension of rural households’ LWI, and it can influence
rural households’ livelihood strategies and decision-making (Dehghani Pour et al., 2018).
Based on the SLF and the literature, five dimensions—human, natural, physical, financial,
and social capital—are selected and are summarized in Table 3.
Human capital refers to the quantity and quality of human resources available in the pro-
cess of sustaining livelihoods. According to the life cycle theory, household composition by
age, including infants, preschool children (Cheng et al., 2021), students (Tan et al., 2022),
elderly members (Cheng et al., 2021), the labor force (Xie et al., 2023), and the mean health
of family members (Kuang et al., 2019), is selected to measure human capital.
Natural capital refers to resources and services derived from natural ecosystems that sup-
port rural livelihoods. In this study, natural capital is characterized by paddy and dryland are-
as contracted by rural households from the rural collective (Kuang et al., 2019), as well as the
perceived quantity and quality of water resources supported by the local ecosystem (Jezeer et
al., 2019).
Physical capital refers to the productive assets and infrastructure used to generate house-
hold income. We select house type (Adjei and Kyei, 2013), energy use (Liu et al., 2022), san-
itation, electrical appliances, entertainment equipment, farm tools, and transportation tools
(Wu et al., 2017) to comprehensively measure rural households’ physical capital.
Financial capital refers to the financial resources that rural households use to achieve their
livelihood goals, which is measured by savings (Jezeer et al., 2019), financial products (Xu et
al., 2018), and annual household income in this study.
Social capital refers to the social resources that rural households acquire through member-
ship in social networks, social organizations, or social groups (DFID, 1999). In this study,
social capital is measured by whether a household has family members, relatives, or friends
working in village collectives and local government, as well as a household’s involvement in
collective affairs, its relationships with villagers, its ease of borrowing and its number of
helpers in times of need (Kuang et al., 2019). Higher social network heterogeneity and com-
munity participation promote social connectedness, allowing households to access more so-
1206 Journal of Geographical Sciences
cial resources and support their livelihood development (Wu et al., 2017).
2.4 Conceptual framework
Thus, based on the abovementioned analyses and our research questions, we established the
conceptual framework of this study, which links the well-being of rural households with its
influencing factors (Figure 1). The proposed pathways are indicated in the framework.
Figure 1 Conceptual framework of rural households’ well-being and its influencing factors
3 Materials and methods
3.1 Study area and local context
The Dabie Mountains are located in East China, at the junction of the Anhui, Hubei, and He-
nan provinces, with the Yellow River to the north and the Yangtze River to the south. The
region is rich in flora and fauna, with national nature reserves in the area; it is also a concen-
trated area of rare and endangered wildlife endemic to China. The China Rural Poverty Alle-
viation and Development Programme (2011–2020) lists the Dabie Mountains as one of the 14
contiguous poverty-stricken areas in China. Although all the poor counties in the Dabie
Mountains have been raised out of poverty, since their productivity development pattern is
dominated by the agricultural economy, the economic structure is somewhat closed in nature
and limited by the complex mountainous geographical environment and the lack of internal
motivation of the rural households themselves; thus, the stability of the region’s poverty alle-
viation needs to be strengthened.
We collected our household samples from two townships in the Dabie Mountains, namely,
Luotuoao township in Luotian county and Shitouzui township in Yinshan county (Figure 2).
Situated at the southern base of the Dabie Mountains, Luotuoao township is renowned for its
diverse agricultural products, including “banli” (Chinese chestnuts), tea, silkworm cocoons,
medicinal herbs, and various other cash crops. With the implementation of China’s rural revi-
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1207
talization strategy, townships have undergone significant socioeconomic changes over the
past decade. This transformation has been largely attributed to the growth of rural ecotourism
and e-commerce facilitated by returnee entrepreneurs. The development of a “3A” tourist at-
traction, namely, the Yan’er Valley Scenic Area, is a notable achievement. These initiatives
have significantly enhanced the number of employment opportunities for and the income
growth of rural households. Shitouzui township, located near the main peak of the Dabie
Mountains and at relatively high altitudes, boasts unique natural conditions due to its geo-
graphical location. It is known for its well-developed cultivation of chestnuts, tea trees, Chi-
nese herbs, and other specialties. Rural residents derive a significant portion of their income
from the cultivation of Chinese herbs, which include both cultivated and wild varieties.
Wujiashan National Forest Park, which is located in the northern section of the township, also
generates some tourism revenue for rural households.
While the two townships are representative of the Dabie Mountains in terms of natural re-
sources and the influence of China’s poverty alleviation and revitalization policies, they also
illustrate the diverse and intricate nature of rural livelihoods. These areas have transitioned
from being solely dependent on natural resources to a more diversified approach. The liveli-
hoods of rural households in these areas are shaped not only by the abundance of natural re-
sources but also by a myriad of internal and external risks. Hence, examining the influence of
various risks, geographic accessibility, and rural governance on the well-being of rural
households in these two townships could offer valuable insights for formulating rural devel-
opment policies in the Dabie Mountains.
Figure 2 Location of the Dabie Mountains and our case study area
1208 Journal of Geographical Sciences
3.2 Data sources and processing
The data were collected in the summer of 2022 by our research team, comprising three pro-
fessors and fifteen students, in Luotuoao township and Shitouzui township in Huanggang city,
Hubei Province. The total numbers of households in Luotuoao and Shitouzui are approxi-
mately 4200 and 9000, respectively. We adopted a simple random sampling method and con-
sidered the distribution of respondents’ location, age, gender, and wealth when selecting po-
tential interviewees. The interviews were conducted face-to-face, and the location of each
household’s residence was recorded using a GPS receiver. Our questionnaire collected de-
tailed information about household demographics, capital assets, land uses, agricultural pro-
duction and off-farm activities, household income and expenditure, participation in govern-
ment policies, perceived livelihood risks, indicators of LWI, geographic accessibility, evalua-
tion of rural collectives, etc. Each questionnaire took approximately 1.5 hours to complete.
We preprocessed the collected data to ensure its validity, specifically by excluding outliers,
using the mean substitution method for missing values, and applying logarithms to variables
involving money. However, due to the high number of households working in urban areas or
being absent during our visits, we ultimately collected 522 valid household samples, consti-
tuting approximately 4% of the total sample.
Rural households were classified into four categories based on their primary sources of
livelihood, specifically, whether they were primarily farm or nonfarm and whether they
worked locally or opted to migrate for job opportunities. Type 1 consists of pure-farm house-
holds, which are defined as rural households with no members participating in off-farm work.
Type 2 consists of local part-time households, which have no members working outside the
local area but have at least one member who is locally employed or running a business. Type
3 consists of mixed part-time households, which have both local and nonlocal migrant work-
ers. Type 4 consists of migratory part-time households, which have no members who partici-
pate in local off-farm work. The distributions of the Types 1, 2, 3, and 4 household samples
were found to be 105, 80, 80, and 257, respectively.
3.3 Statistical model
We explore the impact of multiple risks, rural governance, and geographic accessibility on
rural households’ LWI using an ordinary least squares (OLS) model. We also examine how
the two-way interactions between rural governance, geographic accessibility, and livelihood
risk and their three-way interactions affect rural households’ LWI. The equations are as fol-
lows:
α
ijiji
yx
β
ε
=+ +
(1)
where yi is the dependent variable—the rural household’s LWI for the i-th household, xij de-
notes the j-th explanatory variable for the i-th household, α is the intercept, εi is the corre-
sponding random error term, and βj is the estimated parameter for the j-th explanatory varia-
ble. There are 38 explanatory variables in this model, including variables related to multiple
risks, geographic accessibility, rural governance, and livelihood capital dimensions.
We also explore various possible interaction effects, including the interaction between
multiple risks and rural governance and geographical accessibility. The two models with in-
teraction terms are referred to here as Model 3 and Model 4. The general equations for Model
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1209
3 and Model 4 can be written as follows:
12
αγ
ijijiii
yxxx
β
ε
=+ + +
(2)
where x1 represents the core explanatory variable of multiple risks, and x2 represents the mod-
erating variables, i.e., rural governance and geographical accessibility. γ represents the coeffi-
cient of the two-way interaction term.
To explore the joint impact of rural governance and geographical accessibility on multiple
risks, we develop a model with a three-way interaction (Model 5), and the general equation is
shown below:
12 13 23 123
αδθ μ
i jij ii ii i i ii i i
y x xx xx x x xx x
β
ϑε
=+ + + + + +
(3)
where x1 represents the core explanatory variable of multiple risks, x2 represents the moderat-
ing variable of rural governance, and x3 represents another moderating variable of geograph-
ical accessibility. δ, θ, and
ϑ
represent the coefficients of the two-way interaction terms, and
μ is the coefficient of the three-way interaction.
4 Results
4.1 Descriptive statistics
4.1.1 Rural household well-being
As shown in Table 1, there is not much variation in the perception scores for the four indica-
tors characterizing rural households’ LWI among the 522 respondents. The indicator of ease
of covering living expenses is the most fragmented, with the most variance (SD=0.909), while
job satisfaction is rated the most consistently, with the least variation in ratings (SD=0.869).
The mean scores for the well-being measures are all at a high level, but there are undeniable
differences between the means, with the highest mean score being for satisfaction with hous-
ing conditions (mean=3.729), the lowest mean score being for ease of bearing living expenses
(mean=3.141), and the scores for satisfaction with life satisfaction (mean=3.617) and job sat-
isfaction (mean=3.457) being in the middle.
The LWI takes values ranging between 0 and 1; the spatial distribution of the LWI (Figure 3)
shows that rural households with higher LWI scores are more likely to be located in areas of
high geographic accessibility and concentration. Some of the rural households with higher
LWI are located along main roads, where they can develop small businesses to diversify their
livelihood options.
4.1.2 Multiple risks
Among the four categories of risk (Table 2), environmental risk is at a low risk level, with the
risk of wildlife crop raiding being the highest and that of crop diseases and insect pests being
the lowest within the environmental risk. Environmental risk is most prevalent type of risk
among rural households that are involved in full-time or part-time farming. In terms of policy
risk, 43.7% of respondents reported participating in PES programs; only 9% reported not being
enrolled in rural social security, and the overall reliance of their livelihoods on government
transfers is low. Among health risks, illness is the greatest risk that reduces family labor and
increases economic burden, which is also a major reason for poverty recurrence. Moreover,
1210 Journal of Geographical Sciences
Figure 3 Spatial distributions of the livelihood well-being index
28 of the 522 rural households surveyed reported having suffered livestock losses, resulting in
considerable economic losses. Notably, 96.6% of the household samples reported being en-
rolled in rural medical insurance. We measure household development risk based on four di-
mensions that are prone to large expenditures. The maximum value of a single expenditure
comes from the construction or renovation of a home. In terms of average annual expenditures,
social gifts account for the largest amount, followed by education and medical expenditures,
while house building and renovation expenditures account for the lowest, as most households
build and renovate their homes only a few times throughout their lives.
4.1.3 Geographic accessibility
The measurement of the various indicators of geographic accessibility shows that there is
wide variation in the accessibility of rural households to various geographical points. Of all
the geographic accessibility measures (Table 2), the average time spent visiting the medical
site is the lowest (mean=13.30). Overall, we find that the accessibility of basic education and
health care in the study area is high in a geographical sense. Differences in accessibility not
only reveal differences in distance traveled but are also the result of a combination of a rural
household’s choice of the level of service needed, differences in the means of access used,
accessibility and distance, and the fact that access to goods and services is largely limited by
geographic accessibility, which is an important method of overcoming geographical con-
straints.
4.1.4 Rural governance
Table 2 shows that rural residents have relatively high evaluations of environmental govern-
ance in their villages, as shown by their ratings of village greenness (mean=3.873) and road
cleanliness (mean=3.603). Access to collective information is also satisfactory, with a mean
value of 3.811. Although resource allocation directly affects rural households’ LWI, we find
that the average score for fairness in resource allocation is only 2.996, which is the lowest
score among all rural governance indicators. It is also necessary to improve cultural and en-
tertainment services as well as employment services.
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1211
4.1.5 Livelihood capital
According to the descriptive statistics on human capital shown in Table 3, the average educa-
tion score for household members is 3.483, which indicates that most households have at least
one member with a high school education or higher. A mean health score greater than 4 indi-
cates that the health of each household member interviewed is mostly in a healthy state.
In terms of natural capital, rural households’ land holdings or contracts vary considerably
regardless of whether they consist of dry land or paddy land, which is highly dependent on the
type of livelihood of the rural household. The average household landholding is smaller, and
the average household holding of paddy land (2.315) is larger than the average household
holding of dry land (1.293).
In terms of physical capital, the mean values for house type, energy use, and sanitation are
4.263, 3.695, and 3.637, respectively; these results suggest that the rural households inter-
viewed are of various types, mostly two-story and above, with a wide variety of fuels, includ-
ing not only traditional fuel wood but also clean energy sources such as gas and electricity,
which are available to rural households, most of whom use a mixture of fuel wood and one or
two types of clean energy. Of all the physical capital measures, the average score for sanita-
tion is the highest when measured uniformly, and the distribution of scores is more concen-
trated, which suggests that rural households have better sanitation facilities.
At the level of financial capital, approximately half of the households interviewed reported
having savings (mean=0.503), only a few households reported having some members insured
by commercial insurance (mean=0.305), and the number of households that reported holding
financial products is even smaller (mean=0.012), which shows that financial capital accumu-
lation is weak in the capital component of rural households.
In terms of social capital, the job alienation score for rural households with high social par-
ticipation is generally low, and the average level of participation in collective affairs is mod-
erate (mean=2.578); however, neighborhood relations are found to be harmonious, and social
morale is good.
4.1.6 Heterogeneity in households
The results of the subgroup descriptive statistics show that rural households exhibit the
strongest between-group heterogeneity in terms of geographic accessibility (Figure 4). Partic-
ularly, between Type 1 and other types of rural households, there are two comparative be-
tween-group differences found in physical capital, policy risk, and geographic accessibility.
There is no significant difference found between Type 2 and Type 3 households. In addition,
Type 1 and Type 3 rural households exhibit differences in financial capital, social capital, and
health risks; differences between Type 3 and Type 4 households are reflected in social capital
and geographic accessibility, while well-being reflects only a small significant difference be-
tween Type 1 and Type 2 households.
4.2 Determinants of rural households’ well-being
4.2.1 Impact of multiple risks
The regression results shown in Table 4 (Model 1) suggest that two internal risks (i.e., health
(β=−0.149, p<0.01) and development (β=−0.103, p<0.01)) have significant negative impacts
on rural households’ LWI, which is consistent with Hypothesis H1. However, impacts from
1212 Journal of Geographical Sciences
Figure 4 Descriptive statistics for Type 1, 2, 3 and 4 households
Note: *p<0.05, **p<0.01, and ***p<0.001.
the two external factors (i.e., policy and environmental risks) are insignificant. The regression
results shown in Table 4 (Model 2) suggest that multiple risks have a significant negative im-
pact on LWI in general (β=−0.219, p<0.01).
Specifically, the LWI of rural households is significantly affected by internal risk factors.
A household’s LWI is heavily influenced by its members’ health status. Sick members may
suffer physical weakness or even die as a result of their illness. A seriously ill person is de-
pendent on other family members to survive, which further reduces a family’s labor and
productivity. In addition, treatment, rehabilitation, and day-to-day care can incur considerable
costs, which can put considerable pressure on the family financially. The lack of medical in-
surance thus invariably increases the potential health risks to households, thus laying the
groundwork for weakening the well-being of rural households. Financial deficits resulting
from large or irregular expenditures on house construction, gifts and other expenses can cause
intrahousehold stress and thus undermine the LWI of rural households.
The impact of external factors on rural households’ LWI is not significant. This is because
the sampled households are located in areas with a good ecological environment and a suita-
ble climate, which causes their daily life to be less affected by environmental risks; addition-
ally, the high degree of part-time employment and off-farming among rural households
weakens the risk of environmental and agriculture-related policy shocks.
4.2.2 Impact of geographic accessibility
As shown in Table 4 (Model 1), geographic accessibility has a significant positive effect on
rural households’ LWI (β=0.064, p<0.1), validating Hypothesis H2. The greater the geo-
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1213
graphic accessibility is, the less time it will take or the lower the costs will be for rural
households to access employment markets (county and township centers) and public services
(primary and secondary schools, medical sites). Specifically, rural households are more likely
to travel to counties or towns in search of employment opportunities if they spend less time
traveling by the most common means of transportation, thereby contributing to the shift from
primarily agricultural to a more rewarding agro-industrial mode of work for rural households.
This would increase their income and reduce their exposure to livelihood risks. Additionally,
the more convenient it is to reach commercial centers, the easier it is for rural households to
meet their livelihood needs, which ultimately improves their quality of life. Having primary
and secondary schools nearby means that students’ commute costs and times are reduced,
while parents’ working hours are not compressed. Additionally, the shorter the distance to the
nearest health care facility is, the easier it is to access health services for family members, and
the more secure their health status is.
4.2.3 Impact of rural governance
As expected, rural governance also has a significant positive impact on rural households’ LWI,
and of all the explanatory variables, rural governance makes the largest contribution (β=0.228,
p<0.01) (Table 4, Model 1); this is consistent with Hypothesis H4. Rural governance bodies
provide services to rural households through the development and orderly operation of rules
and regulations covering all aspects of their lives that directly or indirectly affect their LWI.
The construction of cultural and recreational facilities provides the basis for residents to exer-
cise and is conducive to their physical and mental health. The release of employment infor-
mation helps residents breakdown information barriers and widen employment channels,
providing rural households with more diversified choices and increasing their income, thereby
improving their LWI. The fairer the distribution of resources is and the greater the transpar-
ency of collective affairs is, the more favorable it is for rural households to safeguard the au-
thority of social governance subjects.
4.2.4 Impact of livelihood capital
The regression results shown in Model 1 (Table 4) suggest that natural physical (β=0.204,
p<0.01), financial (β=0.182, p<0.01), and social capital (β=0.150, p<0.01) have significant
positive effects on rural households’ LWI, but the effect of human capital is insignificant. As
a material basis for production, natural capital plays a very important role in rural households’
LWI. There is a greater chance of high agricultural output and greater livelihood security if
there is a larger area of contracted paddy and dryland, a more abundant water supply and bet-
ter water quality, resulting in a higher LWI. Moreover, households with a higher LWI are as-
sociated with better housing conditions and more advanced equipment and facilities, which
can lead to a better quality of life and a greater sense of satisfaction with their lives. Further-
more, the more financial capital a rural household has, the more financially secure its life will
be. Last, households with high levels of social network heterogeneity can leverage their own
social connections to gain access to more off-farm employment opportunities, better educa-
tion, markets, and other resources, which tends to lead to higher levels of job and life satisfac-
tion and thus drives households to maintain high levels of LWI.
1214 Journal of Geographical Sciences
Table 4 Regression results: main effects and interaction effects
Va ri a b l e Coefficient (β)
Model 1 Model 2 Model 3 Model 4 Model 5
Human capital −0.006 −0.007 −0.009 −0.008 −0.014
Natural capital 0.204*** 0.192*** 0.184*** 0.237*** 0.194***
Physical capital 0.098** 0.113
** 0.116** 0.108** 0.105**
Financial capital 0.182*** 0.182*** 0.197*** 0.204*** 0.190***
Social capital 0.150*** 0.142*** 0.144*** 0.176*** 0.145***
Geographic accessibility 0.064* 0.066* 0.065* 0.046
Rural governance 0.228*** 0.234*** 0.232*** 0.222***
Environmental risks 0.027
Health risks −0.149***
Development risks −0.103***
Policy risks −0.044
Multiple risks −0.219*** −0.212*** −0.260*** −0.197***
Two-way interactions
Multiple risks×Rural governance 0.455 0.405
Multiple risks×Geographic accessibility 0.722** 0.482
Geographic accessibility×Rural governance −0.123
Three-way interactions
Multiple risks×Geographic accessibility ×Rural governance −3.011
Constant 0.260*** 0.250*** 0.296*** 0.352*** 0.274***
R2 0.252 0.243 0.241 0.210 0.256
Note: *p<0.10, **p<0.05, and ***p<0.01.
4.3 Moderating role of rural governance and geographic accessibility
The moderating role of rural governance and geographic accessibility in mitigating the im-
pacts of multiple risks on rural households’ LWI is explored using Equation (2). To assist in
the interpretation of their interactions, we calculate and graph the predicted probabilities of
rural households’ LWI based on Equation (2) (Figure 5). The values of the dependent varia-
bles (rural households’ LWI) are predicted by changing the value of the multiple risks from
low (−1.0 SD) to medium (mean) to high (+1.0 SD), while geographic accessibility varies
continuously from low to high (Figure 5a), rural governance varies continuously from low to
high (Figure 5b), and both variables vary continuously from low to high (Figure 5c).
4.3.1 Interaction between geographic accessibility and multiple risks
The interaction results between geographic accessibility and multiple risks shown in Model 4
(Table 4) indicate that there is a strong interaction effect between geographic accessibility and
multiple risk and that the interaction has a significant positive effect on rural households’ LWI.
As depicted in Figure 5a, geographic accessibility can weaken the negative impact of multiple
risks on the LWI of rural households by leveraging the slope. When the value of multiple
risks changes from mean to high, the LWI decreases by 0.048 for low geographic accessibility
and 0.015 for high geographic accessibility. This indicates that geographic accessibility can
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1215
Figure 5 Graphical illustration of interaction effects
mitigate the negative impact of multiple risks on rural households’ well-being, thus verifying
Hypothesis H3.
4.3.2 Interaction between rural governance and multiple risks
The interaction effect between rural governance and livelihood risk on rural households’ LWI
is also positive (Model 3) (Table 4). Figure 5b also shows that the slope of high rural govern-
ance is smaller than that of low rural governance. This indicates that for the same variation in
multiple risks, the rate of negative change in rural households’ LWI is lower under high rural
governance. When the value of multiple risks changes from mean to high, the LWI decreases
by 0.034 under low rural governance, while under high rural governance, the LWI decreases
by only 0.017. However, the interaction term is insignificant. Thus, Hypothesis H5 is not to-
tally supported.
4.3.3 Interaction between rural governance, geographic accessibility, and multiple risks
The results of the OLS regression shown in Model 5 (Table 4) suggest that the three interac-
tions of rural governance, geographic accessibility and multiple risks have an insignificant
negative correlation with rural households’ LWI. This result is also shown in Figure 5c. The
figure shows that rural governance and geographic accessibility can together have a negative
1216 Journal of Geographical Sciences
moderating effect on the negative impact of multiple risks on rural households’ LWI. Howev-
er, it is not the case that higher values of both factors are better. The combination of low rural
governance and low geographic accessibility has the worst moderating effect, while surpris-
ingly, the combination of low rural governance and high geographic accessibility has the most
significant moderating effect (Figure 5c), contrary to our expectation. When the value of risk
changes from mean to high, the LWI decreases by 0.567 for low rural governance and low
geographic accessibility conditions. The LWI decreases by 0.016 for the high rural govern-
ance condition, irrespective of the value taken for geographic accessibility, while it decreases
by only 0.009 for low rural governance but high geographic accessibility conditions. The rea-
son is that geographic accessibility and rural governance have a negative interaction effect,
i.e., trade-offs, on rural households’ LWI. With greater geographic accessibility, rural house-
holds rely less on rural governance. This result demonstrates that rural governance is thus less
effective for these rural households and that geographic accessibility may even weaken the
positive role of rural governance in enhancing rural households’ LWI. Consequently, the in-
teraction between rural governance and geographic accessibility fails to mitigate the impact of
livelihood risks on the LWI; thus, ultimately, rural governance, geographic accessibility, and
livelihood risks have a nonsignificant negative impact on rural households’ LWI.
4.4 Heterogeneous effects among household groups
The result of the regression analysis conducted among different household groups shows
strong heterogeneity in the impacts of different factors on rural households’ LWI. In Table 5,
the interaction between geographic accessibility and multiple risks is highlighted; we can see
that geographic accessibility does not play a significant moderating role in the relationship
between multiple risks and rural households’ LWI for all household groups. Geographic ac-
cessibility has a direct and significant positive impact on Type 4 households (β=0.126, p<0.05)
but a nonsignificant negative impact on the other three types of households.
Table 5 Heterogeneity analysis of the interaction between multiple risks and geographic accessibility
Variable Type 1 Type 2 Type 3 Type 4
Human capital 0.081 −0.201* 0.161 −0.005
Natural capital 0.071 0.207 0.196 0.287***
Physical capital 0.266*** 0.131 0.075 0.005
Financial capital 0.204* 0.423*** 0.063 0.224***
Social capital 0.033 0.146* 0.195*** 0.228***
Multiple risks −0.351*** −0.169 −0.034 −0.271***
Geographic accessibility −0.011 −0.031 −0.047 0.126**
Multiple risks×Geographic accessibility 0.589 1.516 −0.148 0.454
Constant 0.612*** 0.645*** 0.627*** 0.611
***
Observations 105 80 80 257
R2 0.268 0.279 0.208 0.230
Note: *p<0.10, **p<0.05, and ***p<0.01.
Table 6 presents the interaction impact between rural governance and multiple risks and
shows that rural governance has significant positive impacts on all four groups, with the
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1217
greatest impacts being on Type 3 households (β=0.329, p<0.01). However, the moderating
effect of rural governance does not show heterogeneity across groups and instead becomes less
significant. Rural governance has a direct positive effect on the well-being of rural households,
such that the more non-farm households there are, the more pronounced the effect is.
Table 6 Heterogeneity analysis of the interaction between multiple risks and rural governance
Variable Type 1 Type 2 Type 3 Type 4
Human capital 0.045 −0.235** 0.139 0.003
Natural capital 0.028 0.147 0.136 0.237***
Physical capital 0.287*** 0.121 0.116 0.000
Financial capital 0.201* 0.384*** 0.036 0.227***
Social capital 0.013 0.111 0.108* 0.213***
Multiple risks −0.233* −0.084 −0.003 −0.214**
Rural governance 0.209* 0.268** 0.329*** 0.198***
Multiple risks×Rural governance 1.253 −0.648 −1.069 0.688
Constant 0.620*** 0.626*** 0.626*** 0.611
***
Observations 105 80 80 257
R2 0.306 0.328 0.292 0.246
Note: *p<0.10, **p<0.05, and ***p<0.01.
The comoderating effects of rural governance and geographic accessibility on multiple
risks exhibit strong intergroup heterogeneity (Table 7). For Type 1 households, the two mod-
erating variables exert a positive moderating effect (β=−10.030, p<0.05), thereby enhancing
the adverse effects of multiple risks. However, the opposite is true for Type 2 households
(β=30.890, p<0.01). For Type 3 and Type 4 households, the dual moderating variables do not
have significant moderating effects. The direct effects of geographic accessibility and rural
governance on rural households’ LWI exhibit strong intergroup heterogeneity. Geographical
accessibility has a significant positive effect on only Type 4 households (β=0.107, p<0.05),
while rural governance has no significant effect on only Type 1 households, and the greater
the degree of nonfarming is, the greater the effect of rural governance is.
4.5 Robustness test
We conduct a series of robustness tests to prove the reliability of our results. First, outliers can
compromise the robustness of the regression results. Despite the elimination of some outliers
during data processing, their effects may persist due to the subjective nature of tail shrinking,
which involves defining the range for tail reduction. To assess the impact of outliers on the
robustness of the regression model, we apply robust regression (Table 8). The results are quite
consistent with those of the original regression. Second, to ascertain whether the choice of
indicators influences the main and interaction effects of regression, we adjust the measures of
geographic accessibility and rural governance. We add the travel time to the nearest scenic
spot and water system to measure geographic accessibility. To further test the effect of the
number of indicators, we rank the rural governance indicators based on loading factors and
exclude the two lowest-scoring factors from the regression. The results, displayed in Tables 9
1218 Journal of Geographical Sciences
Table 7 Heterogeneity analysis of the interaction between multiple risks and geographic accessibility, rural gov-
ernance
Variable Type 1 Type 2 Type 3 Type 4
Human capital 0.029 −0.215* 0.147 0.002
Natural capital 0.034 0.151 0.153 0.247***
Physical capital 0.281*** 0.013 0.156 −0.048
Financial capital 0.185 0.389*** 0.041 0.210***
Social capital 0.084 0.107 0.113* 0.212***
Multiple risks −0.308** −0.206 0.083 −0.187**
Geographic accessibility −0.069 0.033 −0.022 0.107**
Rural governance 0.159 0.249* 0.287** 0.184***
Multiple risks×Rural governance 0.643 −1.325 −1.694* 0.557
Multiple risks×Geographic accessibility −0.121 −0.245 −0.392 0.206
Geographic accessibility×Rural governance −0.224 0.620 0.416 −0.081
Multiple risks×Geographic accessibility×Rural governance −10.030** 30.890*** 5.365 −2.290
Constant 0.619*** 0.613*** 0.629*** 0.614***
Observations 105 80 80 257
R2 0.357 0.406 0.303 0.267
Note: *p<0.10, **p<0.05, and ***p<0.01.
Table 8 Robust regression: main effects and interaction effects
Va ri a b l e Coefficient (β)
Model 1 Model 2 Model 3 Model 4 Model 5
Human capital 0.010 0.011 0.010 0.012 0.005
Natural capital 0.195*** 0.190*** 0.178*** 0.222*** 0.183***
Physical capital 0.143*** 0.152*** 0.152*** 0.134*** 0.136***
Financial capital 0.181*** 0.179*** 0.193*** 0.200*** 0.192***
Social capital 0.160*** 0.151*** 0.153*** 0.184*** 0.150***
Geographic accessibility 0.063* 0.065* 0.063* 0.052
Rural governance 0.208*** 0.214*** 0.211*** 0.208***
Environmental risks 0.019
Health risks −0.137***
Development risks −0.092**
Policy risks −0.021
Multiple risks −0.194*** −0.196*** −0.222*** −0.182***
Two-way interactions
Multiple risks×Rural governance 0.290 0.362
Multiple risks×Geographic accessibility 0.618** 0.517
Geographic accessibility×Rural governance −0.150
Three-way interactions
Multiple risks×Geographic accessibility×Rural governance −2.964
Constant 0.241*** 0.234*** 0.284*** 0.341*** 0.260***
R2 0.236 0.233 0.227 0.200 0.245
Note: *p<0.10, **p<0.05, and ***p<0.01.
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1219
Table 9 Robustness tests for the impact of geographic accessibility
Va ri a b l e Coefficient (β)
Model 1 Model 2 Model 3 Model 4 Model 5
Human capital −0.006 −0.007 −0.009 −0.009 −0.014
Natural capital 0.204*** 0.192*** 0.184*** 0.237*** 0.194***
Physical capital 0.098** 0.113** 0.116** 0.109** 0.105**
Financial capital 0.182*** 0.181*** 0.197*** 0.204*** 0.190***
Social capital 0.151*** 0.142*** 0.144*** 0.177*** 0.145***
Geographic accessibility 0.065* 0.066* 0.065* 0.047
Rural governance 0.229*** 0.234*** 0.232*** 0.223***
Environmental risks 0.027
Health risks −0.150***
Development risks −0.103**
Policy risks −0.045
Multiple risks −0.219*** −0.212*** −0.261*** −0.198***
Two-way interactions
Multiple risks×Rural governance 0.455 0.413
Multiple risks×Geographic accessibility 0.731** 0.492
Geographic accessibility×Rural governance −0.122
Three-way interactions
Multiple risks×Geographic accessibility×Rural governance −3.045
Constant 0.260*** 0.250*** 0.296*** 0.351*** 0.273***
R2 0.252 0.243 0.241 0.209 0.255
Note: *p<0.10, **p<0.05, and ***p<0.01.
and 10, are not significantly different from those of the main model. Finally, to verify the
model fit, we employ bootstrapping for 100 and 1000 repetitions of sampling. The lower the
root mean square error (RMSE) is, the better a given model is able to “fit” a dataset. As
shown in Table 11, the standard error of the RMSE is quite low, suggesting a satisfactory
model fit.
5 Discussion
5.1 Role of geographic accessibility and rural governance in risk mitigation
In this study, we find a significant positive impact of geographic accessibility on rural house-
holds’ LWI for the full sample and Type 4 households. Our study also confirms that the in-
teraction term of geographic accessibility and multiple risks has a significant positive impact
on rural households’ LWI, demonstrating that geographic accessibility can either mitigate or
offset the adverse effects of livelihood risks on rural households’ LWI. These findings are
consistent with the results of other relevant studies (Liang et al., 2022; Lee and Kim, 2023).
First, geographical context and location directly and indirectly shape households’ livelihoods
(Ma et al., 2023). High geographic accessibility can help rural households develop locational
advantages and facilitate access to economic activities, employment opportunities, and social
1220 Journal of Geographical Sciences
Table 10 Robustness tests for the impact of rural governance
Va ri a b l e Coefficient (β)
Model 1 Model 2 Model 3 Model 4 Model 5
Human capital −0.014 −0.015 −0.016 −0.008 −0.019
Natural capital 0.206*** 0.193*** 0.186*** 0.237*** 0.198***
Physical capital 0.082* 0.097** 0.101** 0.108** 0.090*
Financial capital 0.185*** 0.185*** 0.199*** 0.204*** 0.192***
Social capital 0.145*** 0.137*** 0.139*** 0.176*** 0.140***
Geographic accessibility 0.061* 0.064* 0.065* 0.048
Rural governance 0.199*** 0.203*** 0.202*** 0.193***
Environmental risks 0.030
Health risks −0.150***
Development risks −0.105***
Policy risks −0.051
Multiple risks −0.226*** −0.221*** −0.260*** −0.206***
Two-way interactions
Multiple risks×Rural governance 0.313 0.236
Multiple risks×Geographic accessibility 0.722** 0.447
Geographic accessibility×Rural governance −0.178
Three-way interactions
Multiple risks×Geographic accessibility×Rural governance −1.970
Constant 0.290*** 0.279*** 0.322*** 0.352*** 0.298***
R2 0.255 0.246 0.243 0.210 0.257
Note: *p<0.10, **p<0.05, and ***p<0.01.
Table 11 Model fit test
RMSE Reps=100 S.E. (RMSE) Reps=1000 S.E. (RMSE)
Model 1 0.14857 0.00449 0.00476
Model 2 0.14896 0.00463 0.00481
Model 3 0.14920 0.00454 0.00476
Model 4 0.15226 0.00477 0.00517
Model 5 0.14833 0.00466 0.00477
services (Palacios et al., 2022), which can, on the one hand, diversify rural households’ live-
lihood choices to achieve risk diversification and, on the other hand, promote the accumula-
tion of rural households’ livelihood capital and the optimization of intrahousehold labor
structures, thus improving rural households’ ability to alleviate intrahousehold stress and re-
sist external risk shocks. Second, geographic accessibility breaks the geographical limitation
of rural households’ employment; thus, rural households are not bound to spatially con-
strained agricultural-related work, and nonfarming work is less impacted by external risks
such as the environment and policies. The results of our rural household heterogeneity analy-
sis also show that geographic accessibility has a significant positive effect on Type 4 house-
holds. In addition, geographic accessibility itself encompasses a heterogeneous reflection of
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1221
subjective behaviors, where rural households usually adopt different risk management strate-
gies to avoid risk under uncertainty (Lee and Kim, 2023).
We also find that rural governance has a direct significant effect on the LWI for the full
sample, as well as for each type of rural household. However, it does not have a significant
moderating effect on the relationship between multiple risks and households’ LWI. The pos-
sible reasons for this result are as follows. First, owing to the low degree of autonomy in rural
governance, there is an overreliance on the top-down model of governance, which produces
governance decisions that are partly “ex post”, with a lag, and partly “ex ante”, which may be
detrimental to the short-term interests of some stakeholders at the early stages of their imple-
mentation; however, it takes time for the results of effective governance to be evident from
such measures. Second, the emergence of a rural risk society is essentially a matter of the re-
sponsibility of the subject of governance (Peet, 1997); rural governance involves a large
number of stakeholders, some of whom have trade-offs and not all of whom are beneficiaries.
Third, the relationship between rural governance and multiple livelihood risk is not a simple
linear correlation, and rural governance and risk are not directly correlated, as there may exist
mediating or moderating variables that influence their connection. Consequently, moderation
of the relationship between multiple risks and rural households’ well-being by rural govern-
ance does not exhibit statistical significance.
The results of this study are of great practical significance, especially for poverty eradica-
tion areas in the rural revitalization stage. The empirical evidence clearly indicates that geo-
graphic accessibility significantly contributes to alleviating the geographic poverty trap. This
provides a strategic reference for reinforcing poverty eradication outcomes. The spatial dis-
tribution of various resources, guided by accessibility, can serve as a catalyst to improve rural
household well-being. Rural governance can be used as a means of regulation because it can
reflect the degree of subjective initiative of policy-makers and rural managers; such govern-
ance can also be used as a means to harmonize geographic accessibility and rural governance
policies. The optimal solution to the positive interaction can only be found by the two-way
empowerment of geographic accessibility and rural governance; thus, the two-way interaction
of geographic accessibility and rural governance can maximize the moderating effect of the
negative relationship between livelihood risk and rural households’ well-being.
5.2 Policy implications
The inadequate livelihood stability of rural households in poverty-stricken mountainous areas
and multiple risk shocks will lower the well-being of rural households; thus, it is necessary for
rural grassroots governance bodies to establish a risk early warning and feedback mechanism,
to involve rural households in the governance process, to regularly collect the potential risks
of rural households and give them countermeasures, and to make regular visits to assess the
risk level and effectiveness of governance to dispose of the risks in a timely manner and to
help reduce the negative impacts of the risks on the well-being of rural households. Consider-
ing the heterogeneity of rural households, regardless of the interactions between multiple risks,
geographic accessibility and rural governance, multiple risks consistently have a robust nega-
tive impact on both Type 1 and Type 4 households, which suggests that households relying on
a single livelihood strategy are more vulnerable to risk. To mitigate the negative impact of
multiple risks on the well-being of these households, it is essential to promote livelihood di-
1222 Journal of Geographical Sciences
versification, such as enhancing these households’ access to productive assets, offering them
employment training, developing rural industries and tourism, and enhancing their digital
ability.
Geographical accessibility has a significant positive impact on the well-being of rural
households. Geographic accessibility includes economic, market, transportation, infrastructure,
and other dimensions. Among these factors, transportation accessibility has the most direct
impact on the livelihoods of rural households. It also indirectly affects the accessibility of
other resources and services, which further affects rural households’ well-being. Therefore,
priority should be given to ensuring transportation accessibility in villages, increasing public
transportation stations, and expanding the radius of public transportation. Second, the ad-
vantage of geographic accessibility should be given full play in rural development planning,
and the allocation of social supply services such as medical care and education should take
full account of accessibility and optimize the geographic layout, especially in remote moun-
tainous areas where rural households are dispersed. Regarding the heterogeneity in rural
households, geographic accessibility has only a significant positive impact on Type 4 house-
holds, which indicates that the most connected households also have the highest demand for
geographic accessibility; this suggests that enhancing geographic accessibility can be a key
strategy for overcoming the spatial poverty trap. However, for the other three types of house-
holds, the effects of geographic accessibility are not significant, which indicates that the sub-
jective perception of the relationship between the benefits of geographic accessibility and the
well-being of the household is weak. Therefore, the government can facilitate the transporta-
tion of agricultural and forestry products to the city and the commuting of migrant workers
through the opening of special urban and rural lines to effectively enhance the perception of
geographic accessibility by Type 1, Type 2, and Type 3 households.
Rural governance should consider the heterogeneity of rural households and duly
acknowledge and respect their livelihood preferences to maximize their livelihood autonomy.
It is imperative for decision-makers to provide appropriate policy support and for upper-level
governance bodies to delegate substantial authority to lower-level governance bodies. This is
because grassroots governance bodies, which maintain direct contact with rural households,
possess a superior ability to grasp their actual needs and can respond immediately. Rural
households with insufficient endogenous motivation and resilience are more dependent on the
support provided by governance bodies. Fairness in the distribution of resources or services is
strongly associated with the well-being of rural households; thus, the equitable allocation of
resources and provision of services must be ensured. The magnitude of the impact of rural
governance on rural households’ well-being is contingent upon the degree of nonagricultural
activities and the diversity of livelihoods. Specifically, for Type 1 households, the three-way
interactions exhibit a significant negative effect. This outcome can be ascribed to the high
reliance of pure-farmed households on agriculture, leading to increased exposure to ongoing
or potential external risks. Moreover, Type 1 households exhibit less resilience to internal and
external risks due to their homogeneous livelihood strategies and limited intrinsic motivation.
To improve the well-being of Type 1 households, the initial focus should be on risk aversion,
supplemented by the dual optimization of rural governance and geographic accessibility. For
Type 2 households, geographic accessibility plays a pivotal role in terms of primary and in-
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1223
teraction effects. The two-factor interaction between geographic accessibility and rural gov-
ernance alleviates the negative correlation between multiple risks and well-being and sup-
presses the adverse impact of risks. Therefore, the coupling and coordination thresholds be-
tween the two need further clarification, and risk control measures should be enhanced to en-
sure high-quality stabilization.
5.3 Limitations and prospects
This study examines the impacts of geographic accessibility, rural governance, and multiple
risks on rural households’ LWI in a mountainous area in rural China and then explores the
two-way and three-way interactions of these variables. The results of this study provide new
policy insights into managing livelihood risks and enhancing rural households’ well-being
from a geographic and rural governance perspective. This study may have valuable implica-
tions for stabilizing poverty alleviation in mountainous areas and promoting rural revitaliza-
tion.
However, this study has several limitations. First, the dataset utilized in this study is static
and fails to capture the dynamic characteristics of rural households. This absence of temporal
variation might constrain the applicability of our findings over time. Second, the dataset, the
information for which was primarily collected from two townships in the Dabie Mountains, is
considered relatively small for heterogeneity analysis and may lack sufficient representative-
ness. It is therefore suggested that future research should aim to incorporate larger and more
dynamic datasets to enhance the robustness and generalizability of the findings. Furthermore,
while this study primarily concentrates on the subjective well-being of rural households, a
comprehensive framework that evaluates both subjective and objective well-being could pro-
vide valuable insights for the formulation of rural development policies in mountainous re-
gions.
6 Conclusions
This study constructs a conceptual framework consisting of livelihood capital, multiple risks,
geographic accessibility, rural governance, the Livelihood Well-being Index (LWI), and the
relationships between these factors. We explore the moderating effects of geographic accessi-
bility and rural governance on the relationship between multiple risks and rural households’
well-being. The regression results suggest the following results.
Multiple risks, particularly those related to health and development, have a significant neg-
ative impact on the LWI of rural households surveyed, i.e., at the 1% level. In contrast, live-
lihood capital, geographic accessibility, and rural governance have positive effects on rural
households’ LWI. Geographic accessibility has a significant moderating effect on the rela-
tionship between multiple risks and rural households’ well-being, which can help decrease the
adverse effects of multiple risks on the LWI of rural households. Furthermore, rural govern-
ance exerts a positive influence on the LWI across all four types of rural households, with the
magnitude of the impact being directly proportional to the degree of nonfarm activities. Thus,
it is evident that geographic accessibility and rural governance possess the potential to en-
hance the well-being of rural households. Effectively harnessing this potential has significant
positive implications for enhancing the well-being of rural households.
1224 Journal of Geographical Sciences
References
Adjei O W, Kyei P O, 2013. Linkages between income, housing quality and disease occurrence in rural Ghana.
Journal of Housing and the Built Environment, 28: 35–49.
Alatartseva E, Barysheva G, 2015. Well-being: Subjective and objective aspects. Procedia: Social and Behavioral
Sciences, 166: 36–42.
Alemie T C, Buytaert W, Clark J et al., 2022. Barriers to implementing poverty alleviation through livelihood
strategies: A participatory analysis of farming communities in Ethiopia’s upper Blue Nile basin. Environmental
Science & Policy, 136: 453–466.
Amfo B, Aidoo R, Osei Mensah J et al., 2023. Linkage between working conditions and wellbeing: Insight from
migrant and native farmhands on Ghana’s cocoa farms. Heliyon, 9: e13383.
Arinaminpathy N, McLean A R, Godfray H C J, 2009. Future UK land use policy and the risk of infectious disease
in humans, livestock and wild animals. Land Use Policy, 26: S124–S133.
Bi G H, Yang Q Y, 2023. The spatial production of rural settlements as rural homestays in the context of rural revi-
talization: Evidence from a rural tourism experiment in a Chinese village. Land Use Policy, 128: 106600.
Braga C K V, Loureiro C F G, Pereira R H M, 2023. Evaluating the impact of public transport travel time inaccura-
cy and variability on socio-spatial inequalities in accessibility. Journal of Transport Geography, 109: 103590.
Buonocore C, Carlucci F, Ciciarelli L et al., 2023. Accessibility analysis in spatial planning: A case of special eco-
nomic zones (SEZs) in Campania, southern Italy. Land Use Policy, 132: 106763.
Cai X R, Xu C H, Liang Y Q et al., 2023. Glacier service value and influence on human well-being in Qilian
Mountains. Journal of Geographical Sciences, 33: 2211–2236.
Cheng X, Chen J, Jiang S W et al., 2021. The impact of rural land consolidation on household poverty alleviation:
The moderating effects of human capital endowment. Land Use Policy, 109: 105692.
Christoph B J S I R, 2010. The relation between life satisfaction and the material situation: A re-evaluation using
alternative measures. Social Indicators Research, 98: 475–499.
Cuervo L G, Martínez-Herrera E, Cuervo D et al., 2022. Improving equity using dynamic geographic accessibility
data for urban health services planning. Gaceta Sanitaria, 36: 497–499.
D’Exelle B, Munro A, Verschoor A, 2024. Agricultural investment behaviour and contingency: Experimental evi-
dence from Uganda. World Development, 173: 106427.
Dansou H D, Carrier M, 2023. Decentralization, institutional innovation and governance of inter-territorial relations:
A view from Benin. Cities, 133: 104115.
David McKenzie, Aart Perspectives K J J o E, 2014. Do poverty traps exist? Assessing the evidence. Journal of
Economic Perspectives, 28: 127–148.
Davillas A, Burlinson A, Liu H H, 2022. Getting warmer: Fuel poverty, objective and subjective health and
well-being. Energy Economics, 106: 105794.
Dehghani Pour M, Barati A A, Azadi H et al., 2018. Revealing the role of livelihood assets in livelihood strategies:
Towards enhancing conservation and livelihood development in the Hara Biosphere Reserve, Iran. Ecological
Indicators, 94: 336–347.
Deng W, Zhang H, Zhang S Y et al., 2023. How to identify transitional geospace in mountainous areas? An ap-
proach using a transitional index from the perspective of coupled human and natural systems. Journal of Geo-
graphical Sciences, 33: 1205–1225.
DFID, UK, 1999. DFID: Sustainable Livelihoods Guidance Sheets. London, pp.445.
Easterlin R A, Angelescu L, Zweig J S, 2011. The impact of modern economic growth on urban–rural differences in
subjective well-being. World Development, 39: 2187–2198.
Esparcia J, Abbasi F, 2020. Territorial governance and rural development: Challenge or reality? In: Cejudo E, Na-
varro F eds. Neoendogenous Development in European Rural Areas: Results and Lessons. Switzerland: Springer,
33–60.
Georgios C, Barraí H, 2023. Social innovation in rural governance: A comparative case study across the marginal-
ised rural EU. Journal of Rural Studies, 99: 193–203.
Georgios C, Nikolaos N, Michalis P, 2021. Neo-endogenous rural development: A path toward reviving rural Eu-
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1225
rope. Rural Sociology, 86: 911–937.
Ghazali S, Zibaei M, Azadi H, 2023. Impact of livelihood strategies and capitals on rangeland sustainability and
nomads’ poverty: A counterfactual analysis in Southwest Iran. Ecological Economics, 206: 107738.
Ghazali S, Zibaei M, Keshavarz M, 2022. The effectiveness of livelihood management strategies in mitigating
drought impacts and improving livability of pastoralist households. International Journal of Disaster Risk Re-
duction, 77: 103063.
Guo A J, Wei Y, Zhong F L et al., 2022. How do climate change perception and value cognition affect farmers’
sustainable livelihood capacity? An analysis based on an improved DFID sustainable livelihood framework.
Sustainable Production and Consumption, 33: 636–650.
Halla P, Merino-Saum A, Binder C R, 2022. How to link sustainability assessments with local governance? Con-
necting indicators to institutions and controversies. Environmental Impact Assessment Review, 93: 106741.
Jankowska M M, Lopez-Carr D, Funk C et al., 2012. Climate change and human health: Spatial modeling of water
availability, malnutrition, and livelihoods in Mali, Africa. Applied Geography, 33: 4–15.
Jezeer R E, Verweij P A, Boot R G A et al., 2019. Influence of livelihood assets, experienced shocks and perceived
risks on smallholder coffee farming practices in Peru. Journal of Environmental Management, 242: 496–506.
Kearns A, Whitley E, 2020. Are housing and neighbourhood empowerment beneficial for mental health and well-
being? Evidence from disadvantaged communities experiencing regeneration. SSM: Population Health, 12:
100645.
Kuang F Y, Jin J J, He R et al., 2019. Influence of livelihood capital on adaptation strategies: Evidence from rural
households in Wushen Banner, China. Land Use Policy, 89: 104228.
Kubiszewski I, Zakariyya N, Costanza R, 2018. Objective and subjective indicators of life satisfaction in Australia:
How well do people perceive what supports a good life? Ecological Economics, 154: 361–372.
Kumar S, Mishra A K, Pramanik S et al., 2020. Climate risk, vulnerability and resilience: Supporting livelihood of
smallholders in semiarid India. Land Use Policy, 97: 104729.
Lee J, Kim J, 2023. Social equity analysis of public transit accessibility to healthcare might be erroneous when trav-
el time uncertainty impacts are overlooked. Travel Behaviour and Society, 32: 100588.
Li Y H, Jin Q H, Li A, 2022. Understanding the multidimensional poverty in South Asia. Journal of Geographical
Sciences, 32: 2053–2068.
Liang W Q, Li W Y, 2023. Impact of internet usage on the subjective well-being of urban and rural households:
Evidence from Vietnam. Telecommunications Policy, 47: 102518.
Liang Y T, Li S Q, Zeng J Q et al., 2022. Examining the impact of multidimensional accessibility on regional pov-
erty in Laos. Applied Geography, 148: 102789.
Liu J H, Luo X, Liu X J et al., 2022. Rural residents’ acceptance of clean heating: An extended technology ac-
ceptance model considering rural residents’ livelihood capital and perception of clean heating. Energy and
Buildings, 267: 112154.
Liu X P, Su X F, Wang Y J et al., 2014. Review on spatial poverty and deprivation and its enlightenments to pov-
erty geography studies in China. Arid Land Geography, 37: 144–152. (in Chinese)
Liu Y Q, Zhang F Z, Wu F L et al., 2017. The subjective wellbeing of migrants in Guangzhou, China: The impacts
of the social and physical environment. Cities, 60: 333–342.
Liu Y Q, Zhou J, Cheng H et al., 2023. The role of livelihood assets in livelihood strategy choice from the perspec-
tive of macrofungal conservation in nature reserves on the Qinghai-Tibetan Plateau. Global Ecology and Con-
servation, 44: e02478.
Ma L, Zhang Q, Wästfelt A et al., 2023. Understanding the spatiality of the rural poor’s livelihoods in Northeast
China: Geographical context, location and urban hierarchy. Applied Geography, 152: 102865.
Merlin L A, Jehle U, 2023. Global interest in walking accessibility: A scoping review. Transport Reviews, 43:
1021–1054.
Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being: Synthesis.
Min X Q, 2020. Activation and empowerment: From rural governance to rural revitalization. The Journal of Jiangsu
Administration Institute, (6): 45–52. (in Chinese)
Mitropoulos L, Karolemeas C, Tsigdinos S et al., 2023. A composite index for assessing accessibility in urban areas:
1226 Journal of Geographical Sciences
A case study in Central Athens, Greece. Journal of Transport Geography, 108: 103566.
Moatsos M, Lazopoulos A, 2021. Global poverty: A first estimation of its uncertainty. World Development Perspec-
tives, 22: 100315.
Nanor M A, Poku-Boansi M, Adarkwa K K, 2021. Determinants of subjective wellbeing in rural communities:
Evidence from the Juaben Municipality, Ghana. Cities, 113: 103140.
OECD, 2011. Compendium of OECD Well-being Indicators. OECD, 1–40.
Páez A, Scott D M, Morency C, 2012. Measuring accessibility: positive and normative implementations of various
accessibility indicators. Journal of Transport Geography, 25: 141–153.
Palacios A, Gabosi J, Williams C R et al., 2022. Social vulnerability, exposure to environmental risk factors, and
accessibility of healthcare services: Evidence from 2,000+ informal settlements in Argentina. Social Science &
Medicine, 309: 115290.
Palmer C G, Fry A, Libala N et al., 2022. Engaging society and building participatory governance in a rural land-
scape restoration context. Anthropocene, 37: 100320.
Peet R, 1997. Risk, environment and modernity: Towards a new ecology. Journal of Rural Studies, 13: 477–478.
Pot F J, Koster S, Tillema T, 2023. Perceived accessibility in Dutch rural areas: Bridging the gap with accessibility
based on spatial data. Transport Policy, 138: 170–184.
Qiu X, Jin J J, Zhang C Y et al., 2023. Exploring the effects of livelihood capital on clean cooking of rural house-
holds in the Tibetan region of China. Energy and Buildings, 290: 113098.
Rajeev M, Nagendran P, 2023. Protecting land and livelihood under climate risks: What hinders crop insurance
adoption? Land Use Policy, 131: 106711.
Raza A, Zhong M, Akuh R et al., 2023. Public transport equity with the concept of time-dependent accessibility
using Geostatistics methods, Lorenz curves, and Gini coefficients. Case Studies on Transport Policy, 11:
100956.
Ronkko R, Rutherford S, Sen K, 2022. The impact of the COVID-19 pandemic on the poor: Insights from the
Hrishipara diaries. World Development, 149: 105689.
Roy A, Kumar S, Rahaman M, 2024. Exploring climate change impacts on rural livelihoods and adaptation strate-
gies: Reflections from marginalized communities in India. Environmental Development, 49: 100937.
Shen L Y, Li J Y, Du X Y, 2023. An improved evaluation method for public service efficiency from the perspective
of residents′ perception. Environmental Impact Assessment Review, 102: 107205.
Smith H E, Ryan C M, Vollmer F et al., 2019. Impacts of land use intensification on human wellbeing: Evidence
from rural Mozambique. Global Environmental Change, 59: 101976.
Staub C G, Clarkson G, 2021. Farmer-led participatory extension leads Haitian farmers to anticipate climate-related
risks and adjust livelihood strategies. Journal of Rural Studies, 81: 235–245.
Sun P, Zhou L, Ge D Z et al., 2021. How does spatial governance drive rural development in China’s farming areas?
Habitat International, 109: 102320.
Tan J, Cai D L, Han K F et al., 2022. Understanding peasant household’s land transfer decision-making: A perspec-
tive of financial literacy. Land Use Policy, 119: 106189.
Thapa S, 2010. Effectiveness of crop protection methods against wildlife damage: A case study of two villages at
Bardia National Park, Nepal. Crop Protection, 29: 1297–1304.
Tran D D, Park E, Tuoi H T N et al., 2022. Climate change impacts on rice-based livelihood vulnerability in the
lower Vietnamese Mekong Delta: Empirical evidence from Can Tho City and Tra Vinh Province. Environmental
Technology & Innovation, 28: 102834.
Twumasi M A, Jiang Y S, Ameyaw B et al., 2020. The impact of credit accessibility on rural households clean
cooking energy consumption: The case of Ghana. Energy Reports, 6: 974–983.
van Dülmen C, Šimon M, Klärner A, 2022. Transport poverty meets car dependency: A GPS tracking study of so-
cially disadvantaged groups in European rural peripheries. Journal of Transport Geography, 101: 103351.
Wang C, Wang X, Wang Y F et al., 2023. Spatio-temporal analysis of human wellbeing and its coupling relation-
ship with ecosystem services in Shandong province, China. Journal of Geographical Sciences, 33(2): 392–412.
Wang W J, Zhao X Y, Cao J J et al., 2020a. Barriers and requirements to climate change adaptation of mountainous
rural communities in developing countries: The case of the eastern Qinghai-Tibetan Plateau of China. Land Use
WANG Feifan et al.: How geographic accessibility and rural governance mitigate the impact of multiple risks 1227
Policy, 95: 104354.
Wang Y, Zhang Q, Bilsborrow R et al., 2020b. Effects of payments for ecosystem services programs in China on
rural household labor allocation and land use: Identifying complex pathways. Land Use Policy, 99: 105024.
Wang Y, Zhang Q, Li Q R et al., 2021a. Role of social networks in building household livelihood resilience under
payments for ecosystem services programs in a poor rural community in China. Journal of Rural Studies, 86:
208–225.
Wang Y, Zhang Q, Li Q R et al., 2021b. Role of social networks in building household livelihood resilience under
payments for ecosystem services programs in a poor rural community in China. Journal of Rural Studies, 86:
208–225.
Wang Y T, Yan J Z, Wu Y, 2024. Impact of policy measures on smallholders’ livelihood resilience: Evidence from
Hehuang Valley, Tibetan Plateau. Ecological Indicators, 158: 111351.
Wu L, Li C J, Gao Y, 2022. Regional agricultural cooperatives and subjective wellbeing of rural households in
China. Regional Science Policy and Practice, 3: 12502.
Wu Z L, Li B, Hou Y, 2017. Adaptive choice of livelihood patterns in rural households in a farm-pastoral zone: A
case study in Jungar, Inner Mongolia. Land Use Policy, 62: 361–375.
Xiao Y, Xiao Q, Zhang J D, 2023. Balancing the international benefits and risks associated with implementation of
ecological policy on the Qinghai-Tibet Plateau, China. Gondwana Research, 115: 183–190.
Xie H L, Wu Q, Li X B, 2023. Impact of labor transfer differences on terraced fields abandonment: Evidence from
micro-survey of farmers in the mountainous areas of Hunan, Fujian and Jiangxi. Journal of Geographical Sci-
ences, 33(8): 1702–1724.
Xu D D, Liu E L, Wang X X et al., 2018. Rural households’ livelihood capital, risk perception, and willingness to
purchase earthquake disaster insurance: Evidence from southwestern China. International Journal of Environ-
mental Research and Public Health, 15: 1319.
Xu J B, Song J, Li B C et al., 2020. Do settlements isolation and land use changes affect poverty? Evidence from a
mountainous province of China. Journal of Rural Studies, 76: 163–172.
Yasar A, Nazir S, Tabinda A B et al., 2017. Socio-economic, health and agriculture benefits of rural household
biogas plants in energy scarce developing countries: A case study from Pakistan. Renewable Energy, 108:
19–25.
Ye C, Liu Z M, 2020. Rural-urban co-governance: Multi-scale practice. Science Bulletin, 65: 778–780.
Ye Y X, Koch S F, 2023. Towards accessibility or affordability? Multidimensional energy poverty across the South
African urban–rural divide. Energy Research & Social Science, 97: 103002.
Yin Q Q, Sui X Y, Ye B et al., 2022. What role does land consolidation play in the multi-dimensional rural revital-
ization in China? A research synthesis. Land Use Policy, 120: 106261.
Zeng X Y, Guo S L, Deng X et al., 2021. Livelihood risk and adaptation strategies of farmers in earthquake hazard
threatened areas: Evidence from sichuan province, China. International Journal of Disaster Risk Reduction, 53:
101971.
Zeweld W, Van Huylenbroeck G, Tesfay G et al., 2020. Sustainable agricultural practices, environmental risk miti-
gation and livelihood improvements: Empirical evidence from northern Ethiopia. Land Use Policy, 95: 103799.
Zhang H, Zhou B B, Liu S W et al., 2023a. Enhancing intercity transportation will improve the equitable distribu-
tion of high-quality health care in China. Applied Geography, 152: 102892.
Zhang Q Y, Ye C, Duan J J, 2022. Multi-dimensional superposition: Rural collaborative governance in Liushe vil-
lage, Suzhou city. Journal of Rural Studies, 96: 141–153.
Zhang W W, 2022. Social capital, income and subjective well-being: Evidence in rural China. Heliyon, 8: e08705.
Zhang Y C, Torre A, Ehrlich M, 2023b. The impact of Chinese government promoted homestead transfer on labor
migration and household’s well-being: A study in three rural areas. Journal of Asian Economics, 86: 101616.
Zhou Y, Liu Y S, 2022. The geography of poverty: Review and research prospects. Journal of Rural Studies, 93:
408–416.
Zikargae M H, Woldearegay A G, Skjerdal T, 2022. Assessing the roles of stakeholders in community projects on
environmental security and livelihood of impoverished rural society: A nongovernmental organization imple-
mentation strategy in focus. Heliyon, 8: e10987.
