The impact of people's perception of livelihood risks on household consumption in China (Humanities and Social Sciences Communications)
Summary
A journal article in Humanities and Social Sciences Communications (2026) 13:52, DOI 10.1057/s41599-025-05774-z, titled The impact of people's perception of livelihood risks on household consumption in China, by Peisong Han, Liang Li, Mengjie Xu, Yujia Tian and Zekun Wang. The abstract states that the study uses data from the 2014, 2016, 2018, and 2020 waves of the China Family Panel Studies. It reports that perceptions of livelihood risks significantly suppress household consumption, reducing consumption levels and narrowing consumption structure, and that the mechanism analysis finds the effect operates by eroding consumer confidence. It also states that households' risk preferences and proactive government interventions moderate the negative impact.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
- Issuer
- Source documents
Cite this document
“The impact of people's perception of livelihood risks on household consumption in China (Humanities and Social Sciences Communications).”
Original at doi.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/the-impact-of-people-s-perception-of-livelihood-risks-on-household-consumption-in-china/
, SHA-256 af08b58d8f2425f6e939ffc9e82f2b875fc15c6baf5641292977b13e05bc9ce4.Full text
ARTICLE
https://doi.org/10.1057/s41599-025-05774-z OPEN
The impact of people’s perception of livelihood
risks on household consumption in China
Peisong Han1, Liang Li2 ✉, Mengjie Xu3, Yujia Tian4 & Zekun Wang1
Household consumption has emerged as a key driver of economic growth in China. This
study empirically investigates whether perceptions of livelihood risks constrain household
1234567890():,;
consumption, utilizing data from the 2014, 2016, 2018, and 2020 waves of the China Family
Panel Studies (CFPS). The results indicate that perceptions of livelihood risks significantly
suppress household consumption, leading to both reduced consumption levels and a nar-
rower consumption structure. A mechanism analysis reveals that perceptions of livelihood
risks reduce consumption by eroding consumer confidence. Furthermore, households’ risk
preferences and proactive government interventions serve as effective moderating factors,
significantly mitigating the negative impact of perceptions of livelihood risks, thus offering
supporting household consumption. This study contributes new evidence to the growing
literature on the economic impacts of risk perceptions and offers valuable insights for pol-
icymakers seeking to enhance livelihood development and stimulate economic growth.
1 Hunan Agricultural University, Changsha, China. 2 Zaozhuang University, Zaozhuang, China. 3 Guizhou Minzu University, Guiyang, China. 4 Sichuan Film and
Television University, Chengdu, China. ✉email: sway209@163.com
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 1
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
T
Introduction
he confluence of the COVID-19 pandemic, international “crowding-out effect,” causing a significant contraction in
political tensions, and escalating trade protectionism has household consumption expenditure. Luo (2004) analyzed urban
precipitated a structural contraction in the global trade residents in China, finding that unemployment risk, income
system. In response, the Chinese government has accelerated the volatility, and the uncertainty of medical and educational
construction of a new development pattern characterized by expenses all suppressed household consumption. Li and Wang
“domestic circulation as the mainstay, and the mutual promotion (2023) focused on environmental quality, observing that air
of domestic and international circulations.” In this strategic pollution prompts protective measures, which, while safeguarding
transformation, activating residents’ consumption potential and health, limits consumption scenarios and capacity, leading to a
expanding domestic demand have become the core engines of significant reduction in total consumption. Huang et al. (2023)
economic growth. Despite the government’s sustained efforts to argued that stronger earthquake risk perception reduces spend-
implement consumption-stimulus policies, the latent consump- ing, particularly on food, transportation, communication, and
tion potential of Chinese households remains underutilized. Data education/cultural entertainment, with information dissemina-
shows that China’s final consumption expenditure contributes far tion amplifying risk perception and driving consumption
less to GDP growth than that of developed countries such as the declines. Furthermore, factors such as liquidity constraints
United States, the United Kingdom, and Germany, as well as (Hayashi, 1985; Zeldes, 1989), mortgages (Fan and Yavas, 2020),
Asian countries like Japan and South Korea, and even trails digital finance (Chen et al., 2024; Song et al., 2024), household
emerging economies such as India and Brazil (Bai et al., 2012). assets (Eika et al., 2020; Friend and Lieberman, 1975), and
Some scholars attribute China’s “high savings, low consumption” population age structure (Dynan et al., 2009; Han et al., 2022)
dilemma to factors such as liquidity constraints, precautionary significantly influence household consumption. However, various
savings, life cycle and income distribution (Modigliani and Cao, livelihood risks are not isolated; they are interconnected and
2004). These theories explain the impact of persistent economic mutually reinforcing. Whether the comprehensive nature of
uncertainty on consumption but ignore the consequences of perceived livelihood risk exerts a more pronounced suppressive
inadequate social security systems. effect on household consumption remains underexplored.
During this critical period of social transformation, China’s In summary, while existing research confirms that PPLR sig-
social security system remains fraught with shortcomings, nificantly suppresses household consumption, several limitations
exposing residents to a range of livelihood-related risks. For remain: (1) previous research typically employs macro-level
instance, sluggish healthcare reforms have led to costly and proxy measures (e.g., unemployment rates, the Gini coefficient) to
inaccessible medical care. The imbalanced allocation of educa- gauge PPLR; however, these metrics fail to capture the nuanced
tional resources has given rise to problems such as limited school risk expectations that individuals hold regarding education,
choice and high tuition costs. The multi-track pension system has healthcare, pensions, and income inequality; (2) most studies
intensified intergenerational pressures on pension reserves. focus on the impact of a single dimension of PPLR on con-
Moreover, shortages in eldercare and childcare services have sumption, overlooking the systemic complexity of livelihood
sparked conflicts between family responsibilities and work, while service development. Imbalances within the system can give rise
speculative housing market trends have made homeownership to a “barrel effect,” potentially skewing estimates; and (3) despite
increasingly unattainable. Structural adjustments in the labor widespread recognition of the “reservoir effect” of savings on
market have increased employment instability. These challenges consumption, little research explores how PPLR, via modifica-
shape residents’ “risk perception-precautionary savings” mental tions in psychological accounts, triggers reductions in
accounts, thereby affecting their consumption decisions. People’s consumption.
perception of livelihood risks (PPLR) is defined as the subjective This paper, based on data from the 2014–2020 China Family
judgment and expectation formed by residents, based on their Panel Studies, empirically analyzes the impact of PPLR on
personal living conditions and external environmental changes, household consumption and elucidates its underlying mechan-
about the potential economic pressures and uncertainties in key isms. The potential contributions of this study are as three-fold:
livelihood areas such as education, healthcare, housing, employ- (1) by constructing a comprehensive PPLR index through the
ment, and pensions1. Compared to the extreme, low-probability, entropy weight method and empirically testing its consumption-
high-damage events like earthquakes or terrorist attacks, liveli- suppressing effect using micro-level resident survey data, this
hood risks are intrinsically linked to daily life, leading to a much research offers a novel perspective and comprehensive evidence
stronger PPLR and a correspondingly greater impact on con- to explain China’s phenomenon of “high savings and low con-
sumption behavior. sumption”; (2) it uncovers the transmission pathways through
Scholars have revealed the multifaceted impact of livelihood which PPLR affects consumption behavior from three key per-
risk perception on household consumption. Campos and Reggio spectives: consumption confidence, risk preference, and govern-
(2015) analyzed households with employed primary earners, ment intervention; and (3) through the analysis of differences
demonstrating that even the mere possibility of unemployment across various groups and regions, this paper confirms the
significantly reduces total household consumption expenditure. “inverted effect” of risk and wealth distribution in the relationship
Harmenberg and Öberg (2021) differentiated consumption between PPLR and household consumption. While lower social
structures, finding that expectations of unemployment risk pri- strata exhibit greater PPLR intensity, the suppressive effect of
marily drive declines in durable goods consumption, while actual PPLR on consumption behavior remains relatively consistent
unemployment more strongly affects non-durable goods con- across social strata. This finding lends strong support to the core
sumption. Krueger et al. (2006) showed that the widening income proposition of risk society theory, particularly the idea of the
disparity not only reduces total social consumption but also “democratization of risk distribution.”
exacerbates consumption stratification through wealth distribu-
tion effects, leading to a vicious cycle of “wealth gap-consumption
suppression.” Dong et al. (2017), Luo et al. (2004), studied the Theoretical analysis and research hypothesis
housing market, noting that rising housing prices had a non- Theoretical basis. The precautionary savings theory reveals
linear effect on consumption: when price increases exceed residents’ rational strategy to mitigate risks by increasing pre-
affordability, housing assets shift from the “wealth effect” to the cautionary savings and reducing current consumption in
2 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
response to potential income declines or expenditure surges As a result, when residents lose confidence in economic prospects,
(Leland, 1968; Deaton, 1991). This theory integrates elements their PPLR increases significantly, giving rise to a “self-fulfilling
such as risk perception, risk attitude, consumption confidence, pessimistic expectation.” Even absent an immediate economic crisis,
and consumption behavior. Among these, risk perception exerts apprehensions about future economic conditions drive households
the most direct influence on consumption decisions. Preparing to increase precautionary savings, thereby curtailing immediate
for uncertainties in future income and expenditure involves a consumption (Brüggen et al.,2017;Hampson et al.,2021). Vanlaer
cognitive assessment of potential risks. When residents choose to et al. (2020) demonstrated that in the wake of the global financial
increase savings and reduce consumption, this reflects both a crisis, declining consumer confidence significantly amplified
rational decision to maximize utility over time and a manifesta- household savings across 18 EU countries. On this basis, we propose
tion of declining consumption confidence and varying risk atti- Hypothesis 2.
tudes (Andersen et al., 2022). Typically, groups with lower H2: PPLR suppresses household consumption by under-
consumption confidence exhibit more significant reductions in mining consumption confidence.
consumption under risk shocks (Mynaříková and Pošta, 2023).
Risk perception is not isolated but is influenced by individual risk In consumption decision-making, risk attitude serves as a
attitudes. Risk-tolerant individuals perceive potential risks less critical determinant of consumer behavior. Defined as the
acutely, save less, and display more robust consumption behavior willingness of residents to accept risk under uncertainty (Wauters
(Xie et al., 2022), while risk-averse individuals amplify risk per- et al., 2014), risk attitude categorizes individuals into three
ceptions, prioritizing precautionary savings and adopting more groups: risk-seeking, risk-neutral, and risk-averse. Research has
cautious consumption patterns (Giglio et al., 2021). Livelihood shown that consumption decisions are influenced by both risk
risks, rooted in public policies and economic cycles, surpass the perception and risk attitude (Pratt, 1978; Lopes, 1987). Risk
coping capacity of individuals or households, so there is an urgent attitude not only influences the selection of risk management
need for government intervention to distribute risks socially and strategies (Barberis, 2013) but also modulates the intensity of risk
address them at their source through institutional reforms and tar- perception via cognitive filtering mechanisms. Consequently,
geted measures. Therefore, this paper constructs a theoretical model variations in risk attitude give rise to differing levels of risk
of residents’ risk-perceived consumption decision-making, analyzing perception, which in turn lead to distinct consumption behaviors
the correlation among risk perception, consumption confidence, risk (Volland, 2017; Xie et al., 2022). Specifically, risk-averse
attitude, government intervention and consumption behavior. individuals tend to overestimate the likelihood and severity of
potential losses, leading them to significantly reduce consumption
to build a financial safety net (Kai-Ineman and Tversky, 1979).
The direct impact of perceived livelihood risks on consump- Conversely, risk-neutral individuals align their consumption
tion. Empirical research indicates that, in anticipation of sub- primarily with household needs and financial constraints, while
stantial fixed expenditures - such as those for education, medical risk-seeking individuals exhibit greater tolerance for uncertainty -
care, and housing - households tend to adopt pronounced pre- often perceiving it as an opportunity - and may resort to
cautionary saving behaviors by dynamically adjusting their con- borrowing or other strategies to meet future needs, resulting in a
sumption budgets to achieve intertemporal smoothing (De Nardi, lower tendency toward precautionary savings (Guiso and
2010; Morisset and Revoredo 1995; Chen et al., 2020). Further- Paiella,2008). Thus, we propose Hypothesis 3:
more, the direct threat of unemployment risks and anticipated H3: Risk preference mitigates the suppressive effect of PPLR
income instability prompt households to curtail non-essential on household consumption.
expenses (Stephens, 2004; Benito, 2006). In parallel, the widening
According to social risk-buffering theory, government inter-
income gap intensifies consumption segmentation (Aguiar and
ventions can effectively mitigate residents’ economic vulnerability
Bils, 2015), inhibiting consumption upgrading among low-
and stimulate consumption growth through policies such as the
income groups (Lee, 2022). It is important to recognize that
expansion of public services, the development of a robust social
risk comprises both objective threats and subjective perceptions
security system, and fiscal transfer payments (Cameron and Shah,
(Hansson, 2010). Information asymmetry often magnifies risk
2015). Enhancing the provision of basic public services is essential
perception, leading to anxiety and panic that drive overly cautious
for improving livelihoods. Such services yield both “income-
behavior, distorting consumption patterns (Huang, 2023). As
enhancing” (Ganong et al., 2020; Bianchi et al., 2016; Thaler,
PPLR escalates, households increasingly prioritize rigid budget for
1990) and “expenditure-reducing” (Xiong et al., 2022; Yu et al.,
basic necessities - such as food, housing, and education (Li et al.,
2024) effects. These effects elevate residents’ sense of security,
2020) - while cutting back on discretionary spending on luxury
curb the need for precautionary savings, and unleash consump-
goods and entertainment services (Huang, 2023). In light of these
tion potential. Specifically, investments in primary education and
observations, we propose Hypothesis 1:
universal healthcare effectively address limited school access and
H1: PPLR significantly suppresses household consumption.
high medical costs (Liang et al., 2024), promoting human capital
development and increasing household income. In addition,
The mechanism of perceived livelihood risks on consumption. policies such as strict regulations on housing speculation, housing
Consumer confidence, as an affective dimension of PPLR, serves as a renovation subsidies, and affordable housing initiatives mitigate
critical trigger for precautionary saving behavior. Widely recognized the crowding-out effect on other consumption categories (Hui
as a vital indicator of residents’ consumption propensity (Dees, et al., 2018). Social insurance, through its risk-sharing mechan-
2013) and broader economic trends (Bovi, 2009), consumer con- ism, mitigates extreme economic risks faced by residents
fidence exhibits a well-documented positive correlation with con- (Browning and Lusardi, 1996; Hubbard, 1995). Empirical findings
sumption expenditure (Mynaříková and Pošta, 2023). Katona (1968) reveal that once social insurance coverage reaches a critical
noted that optimistic sentiments stimulate spending, while increased threshold, residents’ risk discount factors decline significantly,
pessimism promotes precautionary savings. Furthermore, amplified thereby diminishing the suppressive impact of risk perception on
perception of risks and uncertainty intensifies fear and anxiety (Mi consumption (Engen and Gruber, 2001). Moreover, by system-
et al., 2019; Zheng et al., 2019), prompting an even more pessimistic atically communicating livelihood policies and current develop-
outlook and overestimation of the potential adverse impacts of risk. mental realities, the government can foster a more rational
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 3
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
understanding of livelihood risks among residents, counteracting per the National Bureau of Statistics of China: survival con-
irrational fears that further curtail consumption. On this basis, we sumption (food, clothing, housing), development consumption
propose Hypothesis 4: (healthcare and education & culture), and enjoyment consump-
H4: Government intervention mitigates the suppressive tion (transportation and communication, household equipment
effect of PPLR on household consumption. and supplies, and others). Total household expenditure is calcu-
lated as the sum of these categories, with its natural logarithm as a
proxy for the household consumption level. Drawing on previous
Data and empirical strategies
literature (Qi and Liu, 2020; Hu et al., 2023), the share of
Data. This study empirically investigates the impact of PPLR on
development and enjoyment consumption expenditures relative
household consumption behavior using micro-level data from
to total consumption expenditure is used as a proxy for con-
four waves (2014, 2016, 2018, and 2020) of the China Family
sumption upgrading.
Panel Studies (CFPS). Administered by the Institute of Social
Science Survey at Peking University, the CFPS continuously
People’s perception of livelihood risks. PPLR refers to the residents’
tracks and collect data on household economic behavior and risk
subjective and integrated assessment of both the likelihood and
perception across 25 provincial-level administrative regions in
potential losses associated with uncertain events in the livelihood
mainland China (excluding Xinjiang, Tibet, Qinghai, Inner
sector (e.g., unemployment or sudden increases in medical
Mongolia, Ningxia, and Hainan). This dataset provides a robust
expenses). In alignment with China’s current livelihood policies,
micro-level empirical basis for examining the long-term dynamics
this study utilizes a core item from the CFPS questionnaire -
of consumption. The selection of the 2014–2020 period is justified
“How serious do you think the following issues are in our
for several reasons: (1) this timeframe spans the conclusion of the
country?” - as the measurement instrument. This item probes
12th Five-Year Plan and the implementation of the 13th Five-
perceptions of key livelihood challenges, including employment,
Year Plan, enabling effective observation of the structural evo-
education, housing, healthcare, income disparity, and social
lution in PPLR in the context of national strategic initiatives such
security, rated on a 0–10 Likert scale (0 = not serious at all, 10 =
as targeted poverty alleviation and rural revitalization; (2) the
extremely serious), where higher scores indicate greater perceived
biennial tracking design facilitates the identification of long-term
severity of risk in each domain. By employing the entropy weight
risk accumulation trends, such as employment adjustments fol-
method to assign weights to these dimensions, a comprehensive,
lowing the 2016 supply-side structural reform and the repercus-
multi-dimensional PPLR index is constructed.
sions of the 2018 US-China trade dispute on China’s export-
driven economy - while minimizing distortions from short-term
Consumer confidence. Due to data constraints, this study uses the
volatility inherent in annual data; and (3) as 2020 marks the
CFPS survey question “How confident are you about your
outbreak of the COVID-19 pandemic, the data capture the abrupt
future?” as the indicator for consumer confidence. This item is
impact on high-risk groups, including informal workers and
rated on a 1–5 Likert scale (1 = no confidence, 5 = very con-
interprovincial migrant laborers. Through a longitudinal comparison
fident), with scores positively correlated with consumption con-
with preceding waves, this study isolates the marginal impacts of
fidence, such that higher scores reflect greater consumer
extreme external shocks on the PPLR-consumption relationship,
confidence.
thereby elucidating the intrinsic connection between PPLR and
household consumption behavior. Moreover, province-level eco-
Risk attitude. Due to the absence of direct measures of risk
nomic indicators, such as per capita GDP and financial loan data,
attitude in the CFPS questionnaire, this study constructs a proxy
are introduced as control variables to account for the influence of
variable for risk attitude following the methodology outlined in
regional developmental disparities on micro-level behavior.
the literature (Chen, 2019). Two questionnaire items were
Before the empirical analysis, were processed as follows: (1) a
selected: “Does your household currently hold any financial
hierarchical linkage was established among the household, adult,
products (such as stocks, funds, government bonds, etc.)?” and
and household member databases using the household identifica-
“What is the total return from financial investments in the past
tion code, creating a unified analytical unit; (2) the household
12 months?” These items are used to infer risk preferences based
member identified as “most familiar with household finances and
on household participation in risky financial investments.
capable of answering financial questions” was designated as the
Households that continue to hold risky financial investments
household head, and samples with household heads younger than
after incurring losses in the previous period are assumed to
16 years were excluded; (3) only records with positive household
exhibit high-risk tolerance and a strong pursuit of risk; these are
income and consumption were retained, eliminating cases in
classified as extremely risk-seeking and assigned a value of 3. If
which total annual consumption or survival-related expenditures
residents either break even in the previous period and maintain
were zero; (4) all monetary indicators (e.g., income, expenditure,
their investments, or did not invest previously but start investing
assets, and GDP) were adjusted to constant 2013 prices using the
in the current period, they are deemed to have a moderate risk
Chinese Consumer Price Index (CPI) to account for inflation; (5)
acceptance and are classified as comparatively risk-seeking, with
to ensure sample continuity, comparability, and information
a value of 2. Households that continue holding profitable
richness, only those samples that participated in at least three
investments from the previous period or exit the market after
survey waves were included; and (6) records with missing,
experiencing losses demonstrate a conditional openness to risk,
anomalous, or invalid key variables were removed. After these
adjusting strategies based on investment outcomes; these are
steps, an unbalanced panel dataset was constructed, consisting of
categorized as generally risk-seeking and assigned a value of 1.
30,557 households across four time points: 7959 samples in 2014,
Households that have never participated in risky financial
8191 in 2016, 8146 in 2018, and 6281 in 2020.
markets - precluding determination of whether they are risk-
neutral or risk-averse - are classified as non-risk-seeking and
Variables assigned a value of 0. This classification forms an ordered
Household consumption. This study measures household con- variable ranging from 0 to 3, where higher values indicate
sumption through two dimensions: consumption level and con- greater risk preference, derived from the relationship between
sumption upgrading. The analysis focuses on daily household households’ intertemporal investment decisions and their profit
consumption, with expenditures classified into three categories or loss outcomes.
4 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
Table 1 Definition of variables.
Variable Definition
Gender = 1 if household head is male, and 0 otherwise.
Age age of household head in years.
Married =1 if household head is married, and 0 otherwise.
Healthy The self-rated health of the respondents is 1=unhealthy, 2=average, 3=relatively healthy, 4=very healthy,
5=extremely healthy.
Education years of schooling of household head.
Household size number of household members.
Child dependency ratio The proportion of the population aged 0–14 to the working-age population aged 15–64.
Elderly dependency ratio The proportion of the population aged over 65 to the working-age population aged 15–64.
Urban =1 if household head is married, and 0 otherwise.
Level of economic development Per Capital GDP by Province.
Financial development index Ratio of Year-End Loans Balance of Financial Institutions by Type to GDP.
Table 2 Summary statistics of regression variables.
Variables Min Max Mean Std.Dev
Panel A Dependent variables
Consumption level 688.427 3,889,844 53,984.690 63,113.940
Consumption upgrading 0 1 0.418 0.200
Panel B Variable of interest
Perception of livelihood risks 0 0.150 0.045 0.028
Panel C Instrumental Variable
Community household perception of livelihood risks 0.011 0.095 0.045 0.011
Panel D Mediator Variable
Consumer confidence 1 5 4.039 0.996
Panel E Moderator Variables
Risk preference 0 3 0.156 0.572
Government intervention 1 5 3.455 0.908
Panel F Individual characteristics
Gender 1 2 1.461 0.498
Age 16 95 51.387 13.050
Married 0 1 0.965 0.185
Healthy 1 5 2.861 1.208
Education 0 22 7.685 4.391
Panel G Household characteristics
Household size 1 18 4.237 1.930
Child dependency ratio 0 6 0.289 0.380
Elderly dependency ratio 0 3 0.323 0.573
Urban area 0 1 0.477 0.499
Per capital household income 0 1,988,049.100 20,650.750 53,252.850
Household financial assets 0 14,677,103.718 72,614.080 284,176.900
Panel H Regional Characteristics
Level of economic development 25,795.920 158,487.500 54,134.710 25,951.570
Financial development index 0.753 2.856 1.479 0.489
Government intervention. In this study, the government inter- 2010). Family economic characteristics include household size,
vention variable is constructed using the CFPS survey question, child dependency ratio, elderly dependency ratio, per capita
“How would you rate the overall performance of your county or income, and financial asset holdings (Agarwal and Qian, 2014;
county-level city/district government last year?” This item is Aladangady, 2017; Kukk, 2016). At the regional level, the model
assessed on a 1–5 Likert scale (1 = very dissatisfied, 5 = very controls for macroeconomic indicators including provincial
satisfied) to capture public evaluation of local government per- economic development and financial deepening (Zhang et al.,
formance. The variable indicates the extent of public endorse- 2020). Detailed definitions of these variables are provided in
ment of the local government’s efforts in the preceding year; Table 1.
higher scores suggest effective policy implementation and a Table 2 presents the descriptive statistics of the analyzed
strong sense of accountability in managing livelihood risks. variables. As shown in the data, the sample households have an
average income of 20,650.75 RMB and financial assets of
Control variables. In line with existing literature on household 72,614.08 RMB. Regarding household heads, the majority are
consumption, this paper incorporates control variables across male, with a mean age of 51.39 years, 96% are married, and they
three dimensions: the personal characteristics of the household report average health status and 7.89 years of formal education.
head, family economic characteristics, and regional character- The average household size is 4.24 members. Notably, these
istics. Specifically, personal characteristics of the household head families bear a substantial dependency burden, as evidenced by a
include variables such as gender, age, marital status, health status, combined child-dependency and elderly-dependency ratio of
and education level (Qi and Liu, 2020; Chamon and Prasad, 0.61.
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 5
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
RMB
70000 0.500
60000
0.400
50000
40000 0.300
Consumption upgrading
30000 Consumption level 0.200
20000 Perception of livelihood risks
0.100
10000
0 0.000
2014 2016 2018 2020
Fig. 1 Trends in PPLR and household consumption.
0.050
0.048
0.046
0.044
0.042
0.040
0.038
0.036
Middle income Western Region
Urban
Eastern Region Central Region
High income
Rural
Low income
Low education High education
habitation Income Education Region
Fig. 2 PPRL across different population groups.
Fig. 1 presents the dynamics of PPLR and household welfare. The immense pressure of daily life exacerbates their
consumption among the sample residents from 2014 to 2020. concerns about future uncertainty, creating a stronger risk
The data show significant improvements in both the level and perceptions. It is worth noting that this finding contrasts sharply
structure of household consumption in China. Average annual with the view proposed by Xu and other scholars, who suggested
household consumption rose from 44,639.19 yuan to 61,908.39 that “the higher the social class, the stronger the PPLR” (Xu et al.,
yuan, indicating an overall increase of 38.686%. The highest 2021).
growth rate occurred between 2014 and 2016 (19.831%), followed
by stabilization at 7.498% to 7.662% in subsequent periods. Empirical strategy
Consumption upgrading rose gradually from 0.408 in 2014 to Basic regression model. To examine the impact of PPLR on
0.430 in 2018, but reverted to its initial level in 2020, indicating household consumption, this study constructs the following
temporary obstacles to upgrading. The PPLR index remained low baseline regression model:
overall (ranging from 0.041 to 0.047), though its fluctuations were
pronounced; it increased by 6.818% to 0.047 during 2014–2016,
dropped to 0.041 in 2018, and rebounded to 0.046 in 2020. It is Conit ¼ α0 þ α0 PPLRit þ α1 Pit þ α2 F it þ α3 Rit þ μi þ δt þ εit
noteworthy that in 2020, the slowdown in consumption growth ð1Þ
and a temporary decline in consumption upgrading (–5.116%)
coincided with a 12.195% rebound in PPLR. This phenomenon
suggests that the COVID-19 pandemic may have suppressed Where:
consumption by reinforcing PPLR. - The dependent variable Conit represents the consumption of
Figure 2 highlights a distinct “stratification” pattern in PPLR. household i in year t.
Households in rural and underdeveloped areas of western China, - The independent variable PPLRit denotes the livelihood risk
characterized by fewer years of education and poorer economic perception level of household i in year t.
conditions, exhibit a stronger PPLR, compared to those from - Pit 、F it 、Rit are control variables at the individual, family,
higher social strata. This may be due to their greater exposure to and regional levels, respectively.
risks such as unemployment, employment instability, persistent -α0 、α1 、α2 、α3 are the parameters to be estimated.
income declines, and various challenges in accessing quality - μi represents the province fixed effect.
education, healthcare, improving living conditions, and social - δ t represents the year fixed effect.
6 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
Transmission mechanism model. Consumer confidence sig- Table 3 Baseline regression results.
nificantly influences household consumption levels and con-
sumption upgrading. To clarify the relationship between PPLR,
Variables Consumption level Consumption
consumer confidence, and household consumption and to spe-
upgrading
cifically assess whether consumer confidence mediates the sup-
pressive effect of PPLR on household consumption, this study (1) (2) (3) (4)
adopts the mediation analysis framework proposed by Baron and PPLR −2.029*** −0.720*** −0.322*** −0.131***
Kenny (1986) and Wen et al. (2014). This approach is used to test (0.076) (0.080) (0.042) (0.049)
Hypothesis 2. The specific mediation models are as follows: gender 0.016** 0.005*
(0.004) (0.002)
Con conf it ¼ α þ β0 PPLRit þ β1 Pit þ β2 F it þ β3 Rit þ μi þ δt þ εit age −0.006*** −0.002***
ð2Þ (0.001) (0.001)
Married 0.210*** 0.048***
Conit ¼ α þ β0 PPLRit þ β1 Con conf it þ β2 Pit þ β3 F it þ β4 Rit þ μi þ δ t þ εit (0.010) (0.006)
Healthy −0.005*** −0.016***
ð3Þ (0.002) (0.001)
Where: education 0.014*** 0.002***
(0.001) (0.000)
- Con conf it is the consumer confidence evaluation for
Household size 0.003*** 0.000
household i in year t. (0.001) (0.001)
- β0 、β1 、β2 、β3 are the parameters to be estimated. Child dependency 0.008 0.005
-The meanings of the other variables are the same as in ratio
Model (1). (0.005) (0.003)
Elderly dependency 0.002 0.001
3.4.3 Moderating effect model. In Model (3), the role of consumer ratio
confidence as a potential transmission mechanism in the rela- (0.003) (0.002)
tionship between PPLR and household consumption, with a focus Urban 0.103*** −0.047***
on mitigating future uncertainty risks. However, this analysis has (0.004) (0.003)
limitations, as it does not account for the influences of household Per capita 0.168*** 0.012***
risk tolerance and government intervention measures. To address household income
(0.006) (0.002)
this, we construct a moderation effect model to verify whether
Household financial 0.001 0.000
risk preference and government intervention can effectively
assets
alleviate the suppressive effect of PPLR on household consump- (0.001) (0.001)
tion, thus evaluating Hypothesis 4 and Hypothesis 5. To achieve Level of economic 0.137*** 0.066*
this, we incorporate interaction terms of PPLR and risk pre- development
ference, as well as PPLR and government intervention, into (0.046) (0.030)
Model (1), to construct Moderation Effect Models (4) and (5). Financial −0.039*** 0.009
The specific models are as follows: development index
(0.012) (0.008)
conit ¼ α þ β0 PPLRit þ β1 Risk att it þ β2 PPLRit ´ Risk att it Provincial FE Yes Yes Yes Yes
þβ3 Pit þ β4 F it þ β5 Rit þ μi þ δt þ εit Year FE Yes Yes Yes Yes
Constant 4.937*** 3.719*** 0.407*** 0.119
ð4Þ (0.028) (0.296) (0.018) (0.190)
N 30,577 30,577 30,577 30,577
conit ¼ α þ β0 PPLRit þ β1 Gov inter it þ β2 PPLRit R2 0.128 0.330 0.030 0.059
ð5Þ
´ Gov inter it þ β3 Pit þ β4 F it þ β5 Rit þ μi þ δ t þ εit Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
where:
-Risk att it represents the risk preference of household i in 13.10% downgrade in consumption structure. To illustrate this
year t. impact, the analysis uses the 2020 median household consump-
-Gov inter represents government intervention. tion of 45,170 yuan as a benchmark. According to the results, the
- risk per it ´ Risk att it represents household. average annual consumption per household would decline by
- β0 、β1 、β2 、β3 are the parameters to be estimated. 32,522.40 yuan (45,170 × 72%), with development/enjoyment
-The meanings of the other variables are the same as in expenditure (median proportion of 39.22%) decreasing by 1670.94
Model (1). yuan (32,522.40 × 39.22% × 13.10%). Extrapolating this effect to
China’s 494 million households (per the National Bureau of Statis-
Result tics’ “Seventh National Population Census Bulletin”), national con-
The direct impact of people’s PPLR on household consump- sumption could contract by approximately 160.66 trillion yuan
tion. Table 3, based on the regression results from Model (1), (32,522.40 × 4.94), equivalent to 158.13% of China’s 2020 GDP
indicates that PPLR exerts a significant inhibitory influence on (101.60 trillion yuan). Overall, these results indicate the highly elastic
household consumption. Specifically, Columns (1) and (2) report suppressive effect of PPLR on household consumption, suggesting
the effects of PPLR on household consumption levels without and its potential systemic repercussions for national economic develop-
with control variables, respectively, while Columns (3) and (4) ment, thereby confirming Hypothesis 1.
demonstrate its impact on consumption upgrading under the The regression results in Columns (2) and (4) elucidate the
same conditions. The findings reveal a statistically significant effects of control variables on household consumption. In terms
negative relationship between PPLR and household consumption of the individual characteristics of the household head, the
at the 1% significance level. Each 1-unit increase in PPLR is positive and statistically significant coefficients for gender, marital
associated with a 72% reduction in household consumption and a status, and education level imply that females, married, and more
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 7
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
Table 4 Instrumental Variable Assessment.
Variables PPLR Consumption Consumption Consumption Consumption
Level (IV-2SLS) upgrading (IV-2SLS) Level (IV-LIML) upgrading (IV-LIML)
CHPPLR 0.311***
(0.014)
Corruption of officials 0.004**
(0.000)
PPLR −4.426*** −0.362** −4.427*** −0.363**
(0.594) (0.157) (0.594) (0.158)
Control variables Yes Yes Yes Yes Yes
F Statistics 283.348
(p = 0.000)
Sargan/Anderson-Rubin 0.120 0.029 0.121 0.029
(p = 0.728) (p = 0.818) (p = 0.728) (p = 0.817)
Durbin 42.700 6.647
(p = 0.000) (p = 0.000)
Wu-Hausman 42.699 6.646
(p = 0.000) (p = 0.000)
Provincial FE Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes
Constant −0.019 3.619*** 0.170 3.691*** 0.121
(0.025) (0.361) (0.220) (0.322) (0.196)
N 30,577 30,577 30,577 30,577 30,577
R2 0.326 0.279 0.058 0.279 0.058
educated household heads tend to have higher consumption influenced with factors such as local economic conditions, the
expenditure. This increase is accompanied by a shift in supply of public services, and overall community governance
consumption structure from basic survival-related spending to performance, which are not fully captured by CFPS control
development-oriented and enjoyment-based spending. However, variables, potentially introducing omitted variable bias. Second,
the negative significant coefficients for age and health suggest that individual PPLR is affected by the overall community environ-
older household heads or those in good health tend to exhibit ment while simultaneously influencing the community mean
lower consumption expenditures. through social networks, creating a bidirectional causality issue.
In terms of family characteristics, household size significantly Finally, extreme polarization within a community could distort
influences household consumption, but the analysis reveals no the accuracy of the community average, leading to measurement
notable effects of child dependency ratio or elderly dependency errors.
ratio on consumption outcomes. The urban-rural divide In response, this paper adds the number of corrupt officials2 as
significantly affects household consumption: urban households an additional instrumental variable. The rationale is that
exhibit consumption levels 10.20% higher than those of rural corruption diverts public fiscal resources, exacerbating livelihood
households, yet their consumption upgrading is 4.70% lower. challenges and thereby heightening PPLR, yet it does not directly
This may be due to rural households’ lower costs for essentials affect household consumption. In addition, this paper includes
like food and housing, coupled with rising income levels and livelihood fiscal expenditure as a control variable3. Livelihood
evolving consumption attitudes. Government policies such as fiscal subsidies alleviate the economic pressure in areas such as
“new energy vehicles and home appliances for rural areas” and education, healthcare, and housing (Hu et al., 2025), thereby
the spread of the internet and rural e-commerce have accelerated promoting other types of consumption. Furthermore, regional
the upgrading of rural household consumption structure. livelihood fiscal investment is negatively correlated with risk
Additionally, per capita household income has a positive impact perception and positively correlated with corruption levels (Aizer
on both consumption level and consumption upgrading. et al., 2016; Zhang, 2019).
Finally, there is a significant positive correlation between regional As demonstrated in Table 4, the first-stage regression results
economic development and household consumption. This is (columns (1)–(2)) reveal that community average risk perception
because economic growth generates job opportunities, increases and official corruption both exert a significantly positive influence
household income, and thus enhances consumption capacity. on individual PPLR, confirming the strong explanatory power of
the selected instrumental variables for the endogenous variable
PPLR. Instrument validity tests yield an F-statistic of 283.348
Endogeneity analysis. To address potential endogeneity, this (P = 0.000) and the Sargan test rejects the null hypothesis at the
study adopts the two-stage least squares (2SLS) approach and 1% significance level, thus confirming the exogeneity of the
selects two instrumental variables: the average PPLR of other instruments. In addition, the Durbin and Wu-Hausman statistics
residents in the respondent’s community (excluding the respon- consistently indicate the presence of endogeneity. Furthermore,
dent) and the number of corrupt officials. The selection of the the LIML estimation results presented in columns (4–5) of Table
community average PPLR as an instrument is justified for two 4 closely align with the 2SLS regression outcomes, providing
reasons: (1) residents in the same community share a similar robust evidence that PPLR suppresses household consumption
policy environment, making the community mean a robust proxy and thereby corroborating Hypothesis 1.
for collective PPLR that is strongly correlated with individual
perceptions; and (2) aggregated subjective perceptions at the
community level are unlikely to directly cause differences in Robustness test
household consumption expenditures. Nonetheless, this instru- Replacing the dependent variable. To verify the robustness of the
ment has certain limitations. First, community PPLR is often findings, this study replaces household consumption level with
8 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
Table 5 Robustness checks - replacing the dependent variable.
Variables Engel’s coefficient Consumption Structure Consumption level Consumption upgrading
Upgrading Index
(1) (2) (3) (4)
PPLR 0.086* −0.788** −0.069*** −0.013*
(0.046) (0.378) (0.011) (0.007)
Control variables Yes Yes Yes Yes
Provincial FE Yes Yes Yes Yes
Year FE Yes Yes Yes Yes
Constant 0.582*** −2.192 3.956*** 0.575***
(0.182) (1.499) (0.029) (0.015)
N 30,577 30,577 16693 16693
R2 0.092 0.020 0.276 0.042
Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
Table 6 Robustness checks: propensity score matching regression.
Variable Matching method Treated Controls Difference S.E. T-stat
Consumption level nearest neighbor matching(K = 4) 4.509 4.532 −0.023 0.005 −4.500***
radius matching 4.509 4.531 −0.023 0.005 −4.830***
Kernel Matching 4.508 4.534 −0.026 0.005 −5.580***
Consumption upgrading nearest neighbor matching(K = 4) 0.411 0.418 −0.006 0.003 −2.390**
radius matching 0.411 0.417 −0.005 0.003 −2.150**
Kernel Matching 0.411 0.417 −0.006 0.002 −2.390**
***p < 0.01, **p < 0.05, *p < 0.1.
the Engel coefficient and substitutes the consumption upgrading endogeneity issues in the core explanatory variable, selection bias
measure with a consumption structure advanced index. The may persist due to inherent differences between individuals with
consumption structure advanced index is defined as the ratio of high and low PPLR. To mitigate this bias, we utilize the pro-
development-oriented consumption expenditure to basic con- pensity score matching (PSM) method to balance these differ-
sumption expenditure, where higher values indicate greater ences and verify the robustness of the finding that PPLR
consumption upgrading (Sun, 2023). The regression results, suppresses household consumption. Table 6 presents the average
presented in Columns (1)–(2) of Table 5, demonstrate that the treatment effect on the treated (ATT) values derived from three
effect of perceived livelihood risk (PPLR) on both household matching methods: one-to-four nearest neighbor matching, cali-
consumption level and consumption upgrading remains statisti- per matching, and kernel matching. The results show a significant
cally significant even after these substitutions, confirming the negative effect of PPLR on household consumption across all
robustness of the baseline regression outcomes. matching methods.
Replacement of data sources. In this section, we incorporate data Mechanism analysis
from the Chinese Social Survey (CSS), published by the Institute Transmission mechanism analysis. Consumer confidence, as a key
of Sociology at the Chinese Academy of Social Sciences, to con- indicator of residents’ psychological state, often shapes both
duct a cross-data source robustness test. The CSS database current and long-term consumption decisions. Compared to
comprises mixed cross-sectional data for the years 2015, 2017, optimistic and confident households, pessimistic household tend
2019, and 2021, which, in conjunction with the CFPS data, create to adopt cautious and conservative consumption behavior. The
complementary samples interwoven across cycles. This integra- regression results in Column (1) of Table 7, based on Model (2),
tion provides robust empirical support for studying the con- show that PPLR negatively affects consumer confidence. Further,
sumption effects of PPLR. Variable construction adheres strictly the regression results in Columns (2) and (3), based on Model (3),
to the procedures described in Section “Data”, “Data and Vari- show that the test statistics for consumer confidence are sig-
ables” with the measurement of the dependent variable remaining nificant at the 1% level. These results indicate that PPLR indir-
consistent. The core explanatory variable is derived from the CSS ectly suppresses household consumption by lowering
questionnaire item “Evaluation of the social security provided by consumption confidence, validating Hypothesis 2.
the government” (covering six dimensions: employment, educa-
tion, healthcare, housing, social security, and minimum living Moderation mechanism analysis. As for the moderating role of
guarantee). A PPLR index is constructed using the entropy weight risk attitude, risk attitudes significantly influence household
method. The regression results, presented in Columns (3–4) of consumption behavior, particularly in current consumption
Table 5, show a significantly negative coefficient for PPLR, further decisions. Individuals with greater risk preference tend to
corroborating Hypothesis 1. approach uncertainty proactively, increasing current consump-
tion by reducing their sensitivity to potential risks. This psy-
Propensity score matching regression. Although this study has chological mechanism prompts them to perceive risks as
used instrumental variable methods to address potential opportunities for immediate gratification or potential gains,
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 9
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
Table 7 Potential explanation: consumer confidence. term between PPLR and government intervention has a positive
effect on household consumption levels and consumption
upgrading, with statistical significance at the 5% level for
Variables Consumer Consumption Consumption
consumption levels and the 10% level for consumption upgrad-
confidence level upgrading
ing. These findings indicate the importance of effective govern-
(1) (2) (3) ment intervention in the livelihood sector in mitigating the
PPLR −0.301*** −0.716*** −0.131*** negative effect of PPLR on household consumption. In other
(0.035) (0.079) (0.049) words, robust government performance in addressing livelihood
Consumer 0.015*** 0.002** issues mitigates the negative impact of PPLR on household
confidence
consumption, thereby supporting Hypothesis 4.
(0.002) (0.001)
Control Yes Yes Yes
variables
Provincial FE Yes Yes Yes Heterogeneity analysis
Year FE Yes Yes Yes Heterogeneity analysis across consumption levels and structures.
Constant 3.088*** 3.672*** 0.114 To explore the differential effects of PPLR on consumption
(0.939) (0.296) (0.190) groups, this study stratifies the sample based on the mean values
N 30,577 30,577 30,577 of consumption level and consumption upgrading, and applies
R2 0.082 0.331 0.059 the Chow test to identify group differences. As indicated in Table
Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
9, the Chow test statistics for the consumption level grouping and
the consumption upgrading grouping are 669.11 and 1428.22,
respectively (P < 0.01), indicating significant differences in the
impact of PPLR across consumption groups. Specifically, the
Table 8 Potential explanation: Risk preference and negative impact of PPLR is more pronounced in the high con-
government intervention. sumption level group (β = −0.229) compared to the low con-
sumption level group (β = −0.204). Similarly, PPLR exerts a
greater impact on groups with lower consumption upgrading
Variables consumption level consumption upgrading
(β = −0.143) than on those with higher consumption upgrading
(1) (2) (3) (4) (β = −0.112). With respect to consumption level, high con-
PPLR −0.015*** −0.017*** −0.004*** −0.004*** sumption groups are inclined to allocate a larger income share to
(0.002) (0.002) (0.001) (0.001) high-priced goods or development and enjoyment expenditures,
Risk att 0.048*** 0.020*** which are more susceptible to reduction or substitution as risk
(0.003) (0.003) perceptions intensify, thereby amplifying the effect of PPLR. In
Gov inter 0.006** 0.003** contrast, low consumption groups mainly focus on essential
(0.002) (0.001) items, such as food, clothing, and housing, with limited room for
PPLR×Risk att 0.353*** 0.064 further cuts; therefore, the marginal impact of PPLR is com-
(0.108) (0.090) paratively smaller. Regarding consumption upgrading, groups
PPLR×Gov inter 0.188*** 0.073* with lower upgrading levels, whose consumption patterns are still
(0.067) (0.042)
forming, exhibit high elasticity in demand for development and
Control variables Yes Yes Yes Yes
Provincial FE Yes Yes Yes Yes
enjoyment consumption, leading to greater reductions when risk
Year FE Yes Yes Yes Yes perceptions rise. In contrast, groups with higher consumption
Constant 3.724*** 3.671*** 0.132 0.103 upgrading show more stable patterns, such as regular tourism or
(0.296) (0.296) (0.190) (0.190) advanced healthcare, supported by robust asset allocation and
N 30,577 30,577 30,577 30,577 credit capacity, which buffer short-term liquidity pressures and
R2 0.334 0.330 0.062 0.060 mitigate the adverse influence of PPLR.
Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
Heterogeneity analysis by regional distribution
Urban-rural heterogeneity: The rural consumption market holds
rather than as barriers to consumption. The regression results in vast potential and serves as a critical driver of domestic demand
Columns (1) and (3) of Table 8, derived from Model (5), show (Yu et al., 2024). Subgroup regression results by urban and rural
that the interaction term between PPLR and risk attitudes has a residents (Table 10) show that PPLR has a stronger suppressive
positive effect on household consumption, though its impact on effect on consumption level among urban residents, while its
promoting consumption upgrading is less pronounced. These negative impact on consumption upgrading is more pronounced
findings suggest that risk preference moderates the suppressive among rural residents. Urban residents, characterized by high
impact of PPLR on household consumption, thereby supporting commodification of daily necessities, heavy reliance on social
Hypothesis 3. security, and limited access to substitute resources like land
In terms of the moderating effect of government intervention, (Zhang and Li, 2020), exhibit lower resilience to risk. When PPLR
solving livelihood issues and ensuring public welfare are intensifies, urban residents reduce non-essential service con-
fundamental responsibilities of the government. Whether dealing sumption and opt for lower-priced substitutes, leading to a
with challenges in employment, healthcare, housing, education, marked decline in overall consumption. In contrast, although
income, and social security individually or addressing multiple rising incomes among rural residents drive demand for durable
complex issues in the livelihood sector, the government must goods and leisure services, their development/enjoyment con-
show a high degree of responsibility and proactive engagement. sumption remains highly elastic and less rigid. Consequently,
Inadequate handling of these issues can lead to their entangle- when PPLR increases, the limited wealth accumulation in rural
ment and trigger systemic social risks that ultimately undermine households barely covers basic survival needs, resulting in a sig-
social stability. According to the regression results shown in nificant reduction in development and enjoyment consumption
Columns (2) and (4) of Table 8 from Model (5), the interaction expenditures.
10 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
Table 9 Heterogeneous effects: consumption level and consumption upgrading.
Variables consumption level consumption upgrading
low consumption level High consumption level low Consumption upgrading High consumption upgrading
PPLR −0.204*** −0.229*** −0.143*** −0.112***
(0.068) (0.076) (0.033) (0.029)
Control variables Yes Yes Yes Yes
Provincial FE Yes Yes Yes Yes
Year FE Yes Yes Yes Yes
Constant 3.864*** 4.444*** 0.350*** 0.256
(0.305) (0.272) (0.130) (0.185)
Chow Test 2299.18 3515.29
P-value 0.000 0.000
N 20,456 10,121 16,020 14,557
R2 0.190 0.132 0.049 0.012
Robust standard errors in parentheses；**p < 0.01, **p < 0.05, *p < 0.1.
Table 10 Heterogeneous effects: household registration.
Variables urban household Rural household
consumption level consumption upgrading consumption level consumption upgrading
PPLR −0.741*** −0.167** −0.678*** −0.180***
(0.115) (0.075) (0.109) (0.065)
Control variables Yes Yes Yes Yes
Provincial FE Yes Yes Yes Yes
Year FE Yes Yes Yes Yes
Constant 3.691*** 0.289 3.364*** 0.003
(0.382) (0.256) (0.530) (0.294)
N 14,590 14,590 15,987 15,987
R2 0.315 0.060 0.237 0.048
Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
Heterogeneity by East, Central, and West Regions. In China, consumption levels to a limited extent. Central transfer payments
regional heterogeneity in economic development, livelihood service and informal mutual aid networks among residents provide a
provision, and consumption policy frameworks result in hetero- cushioning effect, mitigating the impact on survival consumption.
geneous impacts of PPLR on household consumption. To explore As a result, the consumption structure remains relatively stable,
these differences, the full sample was divided into three subsamples: exhibiting a pattern of constrained overall consumption with a rigid
Eastern, Central, and Western regions, for separate regression structure, resistant to systemic downgrading.
analyses. Table 11 shows that the suppressive effects of PPLR on
household consumption differ significantly across regions. For Heterogeneity analysis by socioeconomic status
consumption levels, the negative impact is most pronounced in the Heterogeneity by income levels: Income significantly influences
Central region, followed by the Eastern region, and least evident in household consumption. To explore the heterogeneous effects of
the Western region. With respect to consumption upgrading, a PPLR, the sample is stratified into three income groups based on
statistically significant downgrading effect is observed only in the per capita household income: low-income group (per capita
Eastern region, with no significant effects in the Central or Western income ≤15,000 RMB), middle-income group (15,000 RMB < per
regions. This pattern may be attributed to the Eastern region’s capita income ≤35,000 RMB), and high-income group (per capita
service-sector dominance and diverse consumption profile, where income > 35,000 RMB). Table 12 shows that PPLR significantly
development and enjoyment consumption account for an average suppresses consumption levels across all income groups and
of 44.33% of the sample. High market orientation renders this induces a consumption structure downgrade in both low-income
region structurally sensitive to consumption suppression, driving and high-income households. Notably, high-income households
residents to cut non-essential high-end services - such as enter- exhibit the strongest sensitivity, experiencing the greatest reduc-
tainment and tourism - under intense PPLR shocks, giving rise to tion in total consumption and the most pronounced downgrade
consumption downgrading. Conversely, local policy instruments in consumption structure. This may be due to high-income
(consumption vouchers) often subsize everyday necessities and households’ significant allocation of assets to financial invest-
durable goods rather than services, thereby exacerbating structural ments, which are highly sensitive to economic volatility, coupled
contraction. In the Central region, ongoing industrial restructuring with increased precautionary savings driven by high occupational
and rapid urbanization, combined with inadequate social security mobility. Additionally, social status pressures force these house-
and industrial volatility, expose households to income fluctuations holds to cut non-essential spending to maintain their public
and rigid expenditure burdens. This contributes to a broad reduc- image. Consequently, under PPLR conditions, high-income
tion in consumption without differentiation. In the Western region, households face compounded pressures, experiencing a sig-
characterized by lower economic development and a high propor- nificant decline in overall consumption and a structural
tion of survival consumption (60–65% since 2014), PPLR affects downgrade.
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 11
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
Table 11 Heterogeneous effects: regional distribution.
Variables Eastern Region Central Region Western Region
Consumption level Consumption Consumption level Consumption Consumption level Consumption
upgrading upgrading upgrading
PPLR −0.785*** −0.285*** −0.846*** −0.028 −0.543*** −0.040
(0.125) (0.081) (0.146) (0.090) (0.145) (0.084)
Control Yes Yes Yes Yes Yes Yes
variables
Provincial FE Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes
Constant 3.773*** 0.351 3.556*** 0.691 3.357*** −0.130
(0.382) (0.247) (0.945) (0.589) (1.057) (0.616)
N 12,469 12,469 9102 9102 9006 9006
R2 0.406 0.063 0.287 0.053 0.233 0.038
Robust standard errors in parentheses；***p < 0.01, **p < 0.05, *p < 0.1.
Table 12 Heterogeneous effects: income level.
Variables low-income middle-income high-income
consumption level consumption consumption level consumption consumption level consumption
upgrading upgrading upgrading
PPLR −0.783*** −0.155** −0.475*** 0.009 −0.816*** −0.467***
(0.104) (0.062) (0.140) (0.089) (0.235) (0.171)
Control Yes Yes Yes Yes Yes Yes
variables
Provincial FE Yes Yes Yes Yes Yes Yes
Year FE Yes Yes Yes Yes Yes Yes
Constant 4.156*** 0.336 3.700*** -0.362 1.660** 0.065
(0.432) (0.276) (0.501) (0.331) (0.805) (0.536)
N 17,822 17,822 8860 8860 3895 3895
R2 0.199 0.067 0.194 0.068 0.247 0.070
Robust standard errors in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1.
Table 13 Heterogeneous effects: education level.
Variables low-education high-education
consumption level consumption upgrading consumption level consumption upgrading
PPLR −0.758*** −0.164*** −0.557*** −0.421***
(0.092) (0.055) (0.157) (0.112)
Control variables Yes Yes Yes Yes
Provincial FE Yes Yes Yes Yes
Year FE Yes Yes Yes Yes
Constant 3.781*** −0.077 3.634*** 0.548
(0.355) (0.224) (0.526) (0.355)
N 23,145 23,145 7432 7432
R2 0.258 0.068 0.342 0.068
Robust standard errors in parentheses;***p < 0.01, **p < 0.05, *p < 0.1.
Heterogeneity by education level: Education level reflects dif- exerts a greater inhibitive effect on consumption structure
ferences in individuals’ learning abilities, risk perception, and upgrading in the high-education group. These findings suggest
consumption attitudes. To investigate the heterogeneous role of that education level moderates responses to livelihood risks,
education as a human capital factor in moderating consumption differentially influencing household consumption levels and
behavior, we categorize the sample into two groups based on structure.
years of education. With an average of 7.68 years of education in
the sample, individuals with 9 years or fewer are classified as the
low-education group, while those with more than 9 years form Conclusion
the high-education group. The subgroup regression results in This paper develops a theoretical framework to systematically
Table 13 show that PPLR has a stronger suppressive effect on examine the effect of PPLR on household consumption and its
the consumption levels of the low-education group, while it underlying mechanisms, empirically testing these propositions
12 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
using panel data from the CFPS. The main findings are as follows. Received: 17 January 2025; Accepted: 18 August 2025;
(1) PPLR significantly suppresses household consumption,
reducing both overall consumption levels and inducing a down-
grading of the consumption structure. Mechanism analysis shows
that this adverse effect is rooted in a lack of consumer confidence:
heightened risk perceptions erode confidence, strengthen pre- Notes
cautionary saving motives, and reduce current consumption 1 In the international, the concept of “Minsheng” is not common; instead, governments
propensity. However, the moderating roles of risk preferences and and academics often use terms such as “livelihood”, “well-being” and “welfare” to
government intervention can alleviate this inhibitory effect; and discuss the protection of citizens’ rights and the boundaries of State responsibility. In
(2) the suppressive effect of PPLR on consumption varies across China, “Minsheng” has a rich and specific connotation, which is a generalization of the
basic existence, quality of life, and social justice of nationals, specifically, people’s
groups and regions. With respect to consumption levels, house- livelihood covers many areas such as employment, education, medical care, housing,
holds with higher consumption, urban households, those in the old age, childcare, public security and environmental protection, etc. The concept of
eastern region, high‑income groups and less educated groups “seven accesses” was raised in the Report to the 19th CPC National Congress in 2017,
experience stronger effects. For consumption structure, house- which emphasized the need to improve people’s lives, address their concerns,
holds with lower consumption, rural households, those in the strengthen areas of weakness through development, and promote social fairness and
eastern region, high‑income groups and more educated groups justice. The seven aspects include people’s access to childcare, education, employment,
medical services, elderly care, housing, and social assistance.
experience more pronounced downgrading. 2 The number of corrupt officials refers to the number of government officials
The findings of this study offer crucial policy implications. First, prosecuted by the procuratorial authorities who have committed crimes of office.
it is imperative to recognize, in an objective manner, the impor- These data are sourced from the Procuratorial Yearbook of China and the work
tance of ensuring and improving livelihoods for economic and reports of the provincial people’s procuratorates.
social development. On one hand, policies should reduce regional 3 Drawing on previous studies (Gu Cheng, 2024; Wang Liejun, 2023), expenditures
imbalances in livelihood service provision. While the Eastern directly related to livelihood—such as education expenditure, social security and
employment expenditure, healthcare expenditure, and housing expenditure—are
region continues to enhance its public services, concerted efforts defined as livelihood fiscal expenditure. Data are obtained from the National Bureau of
are needed to increase livelihood investments in the Central and Statistics of China.
Western regions through optimized resource allocation, the
establishment of a project supply compensation mechanism, and References
systematic promotion of equitable regional development. On the Agarwal S, Qian W (2014) Consumption and debt response to unanticipated
other hand, policy initiatives should improve the livelihood con- income shocks: evidence from a natural experiment in Singapore. Am Econ
ditions of rural areas and for vulnerable populations by enhancing Rev 104(12):4205–4230
access to quality basic education, developing a robust employment Aguiar M, Bils M (2015) Has consumption inequality mirrored income inequality?
Am Econ Rev 105(9):2725–2756
support framework, strengthening healthcare services, improving Aizer A, Eli S, Ferrie J, Lleras-Muney A (2016) The long-run impact of cash
housing security standards, and reinforcing social security net- transfers to poor families. Am Econ Rev 106(4):935–971
works to address gaps in livelihood development. Aladangady A (2017) Housing wealth and consumption: evidence from geo-
Secondly, it is necessary to build a consumption resilience graphically linked microdata. Am Econ Rev 107(11):3415–3446
support system oriented toward livelihoods. In the Eastern region, Andersen AL, Hansen ET, Johannesen N, Sheridan A(2022) Consumer responses
to the COVID‐19 crisis: evidence from bank account transaction data. Scand
where the consumption structures are highly sensitive, it is J Econ 124(4):905–929
recommended to introduce a “development and enjoyment- Bai CE, Ｗu BZ, Jin Y(2012) Effects of old age insurance contributions con-
oriented consumption security plan.” This plan would involve sumption and saving. Soc Sci China 08:48–71
government and market collaboration to share risks associated Barberis NC (2013) Thirty years of prospect theory in economics: a review and
with consumption structure fluctuations, alongside optimizing assessment. J Econ Perspect 27(1):173–196
Baron RM, Kenny DA (1986) The moderator-mediator variable distinction in
consumption voucher distribution by allocating a portion to social psychological research: Conceptual, strategic, and statistical con-
cultural entertainment, tourism, and healthcare. In the Central siderations. J Personal Soc Psychol 51(6):1173–1182
region, a “social security payment deferral and rights retention” Benito A (2006) Does job insecurity affect household consumption? Oxf Econ Pap
policy should be introduced to ease household consumption 58(1):157–181
pressure during industrial transformation. In the Western region, Bianchi F, Melosi L (2016) Modeling the evolution of expectations and uncertainty
in general equilibrium. Int Econ Rev 57(2):717–756
policymakers should increase the share of central transfer pay- Bovi M (2009) Economic versus psychological forecasting. Evidence from con-
ments used for consumption protection, prioritizing coverage of sumer confidence surveys. J Econ Psychol 30(4):563–574
essential survival needs, including food, clothing, and housing, so Browning M, Lusardi A (1996) Household saving: micro theories and micro facts. J
as to enhance the risk resilience of low-income populations. Econ Lit 34(4):1797–1855
In addition, digital finance has been shown to significantly Brüggen EC, Hogreve J, Holmlund M, Kabadayi S, Löfgren M (2017) Financial
well-being: a conceptualization and research agenda. J Bus Res
mitigate livelihood risks and stimulate consumption potential 79:228–237
(Gomber et al., 2017; Ozili 2018; Li et al., 2020). Therefore, pol- Cameron L, Shah M (2015) Risk-taking behavior in the wake of natural disasters. J
icymakers should accelerate the development of digital infra- Hum Resour 50(2):484–515
structure, deepen the integration of digital financial services across Campos RG, Reggio I (2015) Consumption in the shadow of unemployment. Eur
various livelihood sectors, and enhance public awareness and Econ Rev 78:39–54
Chamon MD, Prasad ES (2010) Why are saving rates of urban households in China
proficiency in utilizing digital finance. These efforts will ensure rising? Am Econ J Macroecon 2(1):93–130
that digital finance can benefit a wider range of population. Chen B, Yang X, Zhong N (2020) Housing demand and household saving rates in
china: evidence from a housing reform. J Hous Econ 49:101693
Chen X (2019) Does mental health affect households’ participation in risky
Data availability financial markets? Evidence from China Family Panel Studies. China Econ
Stud 2:27–42
Due to the restrictions outlined in the CFPS Data Use Agreement De Nardi M, French E, Jones JB (2010) Why do the elderly save? The role of
(https://www.isss.pku.edu.cn/cfps/cjwt/fbxg/1379024.htm), the medical expenses. J Polit Econ 118(1):39–75
study is not permitted to share the raw or processed datasets on Deaton A (1991) Saving and liquidity constraints. Econometrica 59(5):1221–1248
external websites. However, the raw CFPS data is publicly avail- Dees S, Brinca PS (2013) Consumer confidence as a predictor of consumption
able on the official CFPS website (https://www.isss.pku.edu.cn/ spending: evidence for the United States and the Euro area. Int Econ
134:1–14
cfps/).
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 13
ARTICLE HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z
Dong Z, Hui EC, Jia S (2017) How does housing price affect consumption in China: Modigliani F, Cao SL (2004) The Chinese saving puzzle and the life-cycle
wealth effect or substitution effect? Cities 64:1–8 hypothesis. J Econ Lit 42(1):145–170
Dynan KE, Edelberg W, Palumbo MG (2009) The effects of population aging on Morisset J, Revoredo C (1995) Savings and education: a life-cycle model applied to
the relationship among aggregate consumption, saving, and income. Am a panel of 74 countries. No. 1504. The World Bank
Econ Rev 99(2):380–386 Mynaříková L, Pošta V (2023) The effect of consumer confidence and subjective
Eika L, Mogstad M, Vestad OL (2020) What can we learn about household con- well-being on consumers’ spending behavior. J Happiness Stud
sumption expenditure from data on income and assets? J Public Econ 24(2):429–453
189:104163 Ozili PK (2018) Impact of digital finance on financial inclusion and stability. Borsa
Engen EM, Gruber J (2001) Unemployment insurance and precautionary saving. J Istanbul Review
Monetary Econ 47(3):545–579 Pratt JW (1978) Risk aversion in the small and in the large. Econometrica 32:59–79
Fan Y, Yavas A (2020) How does mortgage debt affect household consumption? Qi H, Liu Y (2020) Change of population age structure and the upgrading of
Micro evidence from China. Real Estate Econ 48(1):43–88 household consumption: an empirical research based on CFPS data. China
Friend I, Lieberman C (1975) Short-run asset effects on household savings and Popul Resour Environ 30(12):174–184
consumption: the cross-section evidence. Am Econ Rev 65:624–633 Song Y, Gong Y, Song Y, Chen X (2024) Exploring the impact of digital inclusive
Ganong P, Noel P, Vavra J (2020) US unemployment insurance replacement rates finance on consumption volatility: insights from household entrepreneurship
during the pandemic. J Public Econ 191:104273 and income volatility. Technol Forecast Soc Change 200:123179
Giglio S, Maggiori M, Stroebel J, Utkus S (2021) Five facts about beliefs and Stephens Jr M (2004) Job loss expectations, realizations, and household con-
portfolios. Am Econ Rev 111(5):1481–1522 sumption behavior. Rev Econ Stat 86(1):253–269
Gomber P, Koch JA, Siering M (2017) Digital Finance and FinTech: current Thaler RH (1990) Anomalies: saving, fungibility, and mental accounts. J Econ
research and future research directions. J Bus Econ 87:537–580 Perspect 4(1):193–205
Gu C, Zhang JN (2024) Fiscal decentralization, audit supervision and respon- Vanlaer W, Bielen S, Marneffe W (2020) Consumer confidence and household
siveness of Subnational Governments: empirical evidence from fiscal saving behaviors: a cross-country empirical analysis. Soc Indic Res
expenditure of municipal governments to meet basic living needs. Financ 147:677–721
Trade Econ 45(6):21–37 Volland B (2017) The role of risk and trust attitudes in explaining residential
Guiso L, Paiella M (2008) Risk aversion, wealth, and background risk. J Eur Econ energy demand: evidence from the United Kingdom. Ecol Econ 132:14–30
Assoc 6(6):1109–1150 Wang LJ (2023) Scale, characteristics, and trends of China’s expenditure on Peo-
Hampson DP, Gong S, Xie Y (2021) How consumer confidence affects price ple’s Welfare. J Manag World 39(03):62–69
conscious behavior: the roles of financial vulnerability and locus of control. J Wauters E, Van Winsen F, Yann de Mey LL (2014) Risk perception, attitudes
Bus Res 132:693–704 towards risk and risk management: evidence and implications. Agricl Econ
Han X, Wei C, Cao GY (2022) Aging, generational shifts, and energy consumption 60(9):389–405
in urban China. Proc Natl Acad Sci USA 119(37):e2210853119 Wen ZL, Ye BJ (2014) Analyses of mediating effects: the development of methods
Hansson SO (2010) Risk: objective or subjective, facts or values. J Risk Res and models. Adv Psychol Sci 22(05):731–745. in Chinese
13(2):231–238 Xie X, Tong Z, Xu S (2022) Risk attitudes and household consumption behavior:
Harmenberg K, Öberg E (2021) Consumption dynamics under time-varying evidence from China. Front Public Health 10:922690
unemployment risk. J Monetary Econ 118:350–365 Xiong X, Yu X, Wang Y (2022) The impact of basic public services on residents’
Hayashi F (1985) The effect of liquidity constraints on consumption: a cross- consumption in China. Human Soc Sci Commun 9(1):1–8
sectional analysis. Q J Econ 100(1):183–206 Xu YH, Lai DP (2021) On the risk perception in people’s livelihood and urban
Hu DL, Cai X, Li YJ (2025) Does the Fiscal Expenditure Structure Promote residents’ mental health. J East China Norm Univ 53(02):104–116
Residents’ Service Consumption?Res Financial Econ Issues 3:87–100 Yu S, Ma X, Zhan P (2024) Effects of the resident basic medical insurance reform
Hu D, Zhai C, Zhao S (2023) Does digital finance promote household consumption on household consumption in China. China World Econ 32(1):96–129
upgrading? An analysis based on data from the China family panel studies. Zeldes SP (1989) Consumption and liquidity constraints: an empirical investiga-
Econ Model 125:106377 tion. J Political Econ 97(2):305–346
Huang W (2023) Risk perception, online search and consumption distortion. Zhang X, Yang T, Wang C, Wan G (2020) Digital finance and household
China Econ Q Int 3(2):88–102 consumption: theory and evidence from China. Manag World
Hubbard RG, Skinner J, Zeldes SP (1995) Precautionary saving and social insur- 36(11):48–62
ance. J Political Econ 103(2):360–399 Zhang X (2019) Fiscal decentralization and political centralization in China:
Hui EC, Dong Z, Jia S (2018) Housing price, consumption, and financial market: implications for regional inequality. In China’s Rural Economy after WTO.
evidence from urban household data in China. J Urban Plan Dev Routledge, pp 310–324
144(2):06018001 Zheng C, Zhang J, Guo Y, Zhang Y, Qian L (2019) Disruption and reestablishment
Kai-Ineman D, Tversky A (1979) Prospect theory: an analysis of decision under of place attachment after large-scale disasters: the role of perceived risk,
risk. Econometrica 47(2):363–391 negative emotions, and coping. Int J Disaster Risk Reduct 40:101273
Katona G (1968) On the function of behavioral theory and behavioral research in
economics. Am Econ Rev 58(1):146–150
Krueger D, Perri F (2006) Does income inequality lead to consumption inequality? Acknowledgements
Evidence and theory. Rev Econ Stud 73(1):163–193 National Social Science Fund of China (NSSFC) Major Project (20ZDA038).
Kukk M (2016) How did household indebtedness hamper consumption during the
recession? Evidence from micro data. J Comp Econ 44(3):764–786 Competing interests
Lee JH, Kim W (2022) Impact of production-distribution income gap on household The authors declare no competing interests.
consumption and housing prices in the region. J Public Policy Stud (4):23–30
Leland HE (1968) Saving and uncertainty: the precautionary demand for saving. Q
J Econ 82(3):465–473 Ethical approval
Li J, Wu Y, Xiao JJ (2020) The impact of digital finance on household consump- The data employed in this study were obtained from the China Family Panel Studies
tion: evidence from China. Econ Model 86:317–326 (CFPS). The CFPS dataset is a secondary dataset, so ethical approval was obtained by the
Li Z, Sha Y, Song X, Yang K, ZHao K, Jiang Z, Zhang Q (2020) Impact of risk CFPS project team, not by the authors of this paper. The CFPS project and its data
perception on customer purchase behavior: a meta-analysis. J Bus Ind Mark collection procedures were granted ethical approval by the Institutional Review Board of
35(1):76–96 Peking University (IRB PU). The ethical approval number is IRB00001052-14010.
Li S, Wang M (2023) Effects of improved air quality on household consumption:
empirical evidence from China Family Panel Studies. China Popul Resour
Environ 33(10):24–34
Informed consent
The informed consent of CPFS project can be found here:http://www.isss.pku.edu.cn/
Liang HJ, Zhang HL, Xia XL (2024) Study on basic public services, perceived social
cfps/docs/20200615135138318338.pdf.
justice and consumption of rural residents.World. Agriculture 10:115–127
Lopes LL (1987) Between hope and fear: the psychology of risk. Adv Exp Soc
Psychol 20:255–295 Additional information
Luo C (2004) Uncertainty during economic transition and household consumption Correspondence and requests for materials should be addressed to Liang Li.
behavior in urban China. Econ Res J 39(4):100–106. in Chinese
Mi LH, Ryong KJ, Han JS (2019) The effect of perceived risk on negative emotions, Reprints and permission information is available at http://www.nature.com/reprints
trust, and investment intentions: focused on investors and potential investors.
Int J Tour Hospitality Res 33:19–31
14 HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | https://doi.org/10.1057/s41599-025-05774-z ARTICLE
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License,
which permits any non-commercial use, sharing, distribution and reproduction in any
medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if you modified
the licensed material. You do not have permission under this licence to share adapted
material derived from this article or parts of it. The images or other third party material
in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s
Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/
licenses/by-nc-nd/4.0/.
© The Author(s) 2025
HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS | (2026)13:52 | https://doi.org/10.1057/s41599-025-05774-z 15
File and source
- File
- the-impact-of-people-s-perception-of-livelihood-risks-on-household-consumption-in-china.pdf
- Size
- 539,008 bytes
- SHA-256
- af08b58d8f2425f6e939ffc9e82f2b875fc15c6baf5641292977b13e05bc9ce4
- Original
- doi.org