The Poverty of Farmers in a Main Grain-Producing Area in Northeast China

Abstract

Farmers’ poverty has long been of global concern, mainly in poor rather than affluent areas. The goal of this paper is to better understand the range of poverty in the context of regional differentiation and to enrich knowledge on farmers’ poverty in affluent areas and areas with good natural conditions. A questionnaire survey of poor farmers in the major grain-producing area of Changchun, Northeast China was conducted. Farmers’ poverty was studied from income poverty and multidimensional poverty by intertwining qualitative and quantitative methods. The results indicate that low education levels and poor physical health were most prevalent in poor farmers, followed by income poverty and low living standards. Governmental policies and the macroeconomic situation in the agricultural sector, non-agricultural employment, aging, cultivated land, and family size correlated closely with farmers’ poverty. The macro changes in policies and global trade liberalization in the agricultural sector impacted farmers’ income through the prices of agricultural products and subsidies and influenced the effect of cultivated land. For poor farmers, the effect of employment opportunities in villages was more significant than in urban areas. Aging remains a challenge for farmers’ poverty now and in the future.

1. Introduction

Rural poverty is a worldwide problem, even in wealthy areas [1]. The effective elimination of rural poverty is a pressing challenge for the international community and one of the key sustainable development strategy goals proposed by the United Nations. Most of the poor are living in rural areas [2], and farmers form the dominant group of the rural poor in most areas. Poverty studies have always highlighted area types as farmers’ poverty is significantly related to the geographical context, which differs with changes in area type [3,4,5,6,7]. To the best of the authors’ knowledge, the widespread concern and efforts to study farmers’ poverty mainly focused on countries and regions with large numbers of poor people, such as India [8,9,10] and Bangladesh [11,12]. Another important focus is on areas with poor natural conditions, such as arid areas [13,14,15], mountainous areas [16,17], and areas with frequent natural disasters [18]. The same holds true for research in China. China’s population base of the rural poor is large and relevant research mainly focused on poor areas. These areas commonly have harsh natural environments, high altitudes, remote locations, or are ethnic minorities and border areas [19,20]. For areas with poor natural conditions, typically located in Africa, productivity growth in agriculture always effectively reduces farmers’ poverty [21,22,23,24,25,26], and relevant research focused more on direct impacts on farmers’ poverty.

Farmers’ poverty in areas with advanced economies or better natural conditions has received relatively little attention from scholars or the public because rural poverty is less apparent in developed countries [1,27]. US researchers have published much work on poverty from geographic and spatial perspectives, especially with regard to environmental connections [7], spatial patterns [5], and the association between poverty and distance to metropolitan areas [6,28]. However, in Europe, rural poverty has received relatively little attention over recent decades [3]. Existing studies mainly explored the poverty of older people, immigrants, and its connection with social exclusion [29,30,31,32]. Farmers’ poverty in these wealthy areas received little research attention. Farmers in wealthy areas, or areas with good agricultural conditions, face a more complex and changeable situation because of their close relationship with their external environment, such as policies, the market, urbanization, and the social context. How these factors affect farmers’ poverty must be explored further [33]. The effects of society and economic factors may be more important in wealthy areas or areas with rich natural conditions compared with areas with poor natural conditions. Furthermore, in areas where absolute poverty has been eliminated or has become rare, relative poverty can be more likely alleviated or the remaining absolute poverty can be overcome, which would otherwise be difficult to achieve. In this sense, studying farmers’ poverty in wealthy areas or areas with less limited natural conditions is of more long-term significance. Thus, this is greatly needed in the future stage of the poverty-reduction process when absolute poverty is less common and relative poverty is more common.

Based on the concepts of geographical differentiation, this paper explores the poverty of farmers in major grain-producing areas (MGPAs) in a metropolitan area. The specific questions are: (1) What are the main factors that influence the poverty of farmers in MGPAs? (2) What are the differences and similarities between farmers’ poverty in MGPAs and that of other areas? These questions were addressed based on a questionnaire survey of poor farmers in Changchun, which is the capital city of Jilin Province and one of the most important MGPAs in China. In Changchun, farmers benefit from more livable conditions compared with other poor areas in China and the rest of the world [34].

This study (1) complements evidence of farmers’ poverty in areas with relatively advanced economies and better natural conditions, enriching focal area types for farmers’ poverty, (2) enriches the knowledge exploration of place-related specifics of farmers’ poverty in the context of different geographies. These achievements help to gain a more comprehensive understanding of poverty and its diversity.

The remainder of this paper is structured as follows: Section 2 briefly reviews the methods for measuring poverty and regional differences in farmers’ poverty. Section 3 describes the data source and the methods used to measure poverty and its impact factors. In Section 4, the characteristics and impact factors of farmers’ poverty are analyzed. Section 5 and Section 6 offer a discussion and conclusion of the study, respectively.

2. Literature Review

2.1. Poverty Definition and Measurement

Poverty is a complex concept that spans many disciplines and fields, including sociology, economics, geography, and politics. Gilin et al. generally recognized poverty as a living condition, characterized by low income that does not meet the needs of household consumption [35]. Thus, income poverty can be understood as the original interpretation of poverty, and today, this term is widely used around the world. Internationally comparable income poverty is based on the poverty line standard issued by the World Bank. The current standard is USD 1.9 per person per day, but this value varies across different countries according to actual conditions. China’s current rural poverty line is Chinese Yuan (CNY) 2300 per person per year (constant price of 2010, ~USD 1.6 per person per day).

With the development of society, economy, and inequality, the connotation of poverty has evolved from a single income aspect to multiple aspects. The ‘feasible ability’ theory has been recognized as an important theoretical basis for multidimensional poverty. It posits that poverty not only refers to low income but also encompasses a lack of basic functions such as adequate nutrition, basic medical conditions, basic housing security, and certain educational opportunities [36,37]. Based on ‘feasible ability,’ the concept of multidimensional poverty has been widely accepted and enriched. The multidimensional poverty index (MPI) is a core concept that has been developed greatly. The most extensively used index was proposed by Alkire and contains 10 indicators covering three key aspects of health, education, and living standards [38]. This index was used by the United Nations Development Programme (UNDP) to measure multidimensional poverty on a global scale. While the indicators adopted by different countries and regions vary, all are mostly based on the above three aspects, which have thus become the main theoretical framework for understanding multidimensional poverty.

2.2. Regional Differentiation of Farmers’ Poverty

That poverty has different characteristics and is impacted by different factors in different areas is a widely accepted concept. The diversity of rural poverty not only enhances available knowledge but also presents a major challenge for rural poverty research, which complicates cross-national comparisons and theoretical conceptualizations [3]. The concept of regional differentiation, including the natural, economic, and social characteristics employed by many comparative studies, may offer an entry point for appropriately addressing this challenge. The particular assemblages of economic, political, social, and cultural factors that are associated with different rural places are important for studying rural poverty [39].

In arid areas, typically located in Africa, low soil fertility and soil degradation are directly linked to low productivity and rural poverty [15,40]. Similarly, in such vulnerable areas that are particularly sensitive to climate change (e.g., heatwaves, sea level rise, destruction of coastal zones, and water shortages because of drought), farmers faced elevated poverty risks in the past and will remain to be subject to this risk into the future [41,42]. For those poor areas, increasing the agricultural sector is especially important for rural poverty reduction [26], and promising measures include improving soil fertility [15,43], developing irrigation technology [8,23], and agricultural mechanization [22]. Controlling the prices of agricultural products is also effective [44,45]. In the non-agricultural sector, developing the urban economy, providing non-agricultural employment opportunities which can diversify livelihoods [46], improving the infrastructure in rural areas [17], focusing on education [11], and optimizing credit access [25] are all helpful to reduce poverty. The accessibility of urban areas, which offers more employment opportunities, a wider market, and better services, is a determinant in mountainous areas [17,19] and other remote areas that are far from cities or metropolitan areas [13,28]. However, this does not mean that farmers’ poverty does not also exist in wealthy areas despite their better economic and natural conditions; merely, the incidence of poverty is lower [5,47]. Except for natural and geographical factors, farmers’ poverty is also affected by numerous socio-economic factors, including governmental policies, social systems, like caste and tribes in India, demographic changes, infrastructures, wars, and prevalence of diseases [9,10,14,27]. In developed countries, farmers have always been an at-risk group in rural areas, and this risk mainly originated from low income, especially affecting small-scale farms [27,48]. For specific agriculturally developed areas, the processes of globalization and trade liberalization [1], as well as agricultural product prices are particularly important [45]. The aging of farmers imposes a further challenge both for developed and rapidly developing areas [4,30,32,49,50]. However, aging farmers are not always more likely to be poorer as governments may provide support [51].

The effects of regional differentiation on poverty are not only reflected in the large spatial scale between different area types but also the micro-regional perspective both in urban areas and rural areas [52,53]. The local opportunity structure is a related concept that has been used to study rural poverty. However, linking considerations about the structure of local opportunities and rural poverty with the theory of intersectionality has received only a little attention so far [54]. In rural areas, the livelihood strategies of residents highly depend on resources that are closely linked to residents’ specific place of residence within the rural community, including natural resources and the resources of society, such as enterprises and cooperatives [52].

3. Materials and Methods

3.1. Study Area

As one of the MGPAs in China, Changchun is located in the mid-latitude temperate zone of the northern hemisphere, in the hinterland of the Northeastern Chinese Plain (Figure 1), This area has a lower altitude but a higher level of cultivated land area per capita and grain yield than the national average level (Figure 2), This means more fertile soil and better foundation for the agricultural industry in Changchun compared with other poor areas of China and the rest of the world. The dominant crop in Changchun is corn. Since 2007, China has implemented the minimum guaranteed prices policy for corn to increase farmers’ planting enthusiasm toward safeguarding the national food security and protecting farmers’ income. This resulted in a higher corn price in China than in the rest of the world. Driven by economic profit, policy guarantees, and easier planting process, the sown areas and output of corn increased and eventually accounted for about 83% of the total grain produced in Changchun in 2016, which promoted the simplification trend of regional crop structure. With the cancelation of this policy in 2016, the corn price entered a market-oriented acquisition stage, and thus, experienced clear decrease over recent years. Various subsidies from the government, including subsidies for agriculture producers, fertility protection subsidies for arable land, and social security subsidies of minimum living standard, disability, and family planning have become important to cushion the price impacts on farmers’ income. Changchun is the capital city of Jilin Province and a metropolitan area. The residents’ disposable income per capita is very close to the national average level, and the Gross Domestic Product (GDP) per capita is significantly higher than the national average level (Figure 2). This means that this area has a higher economic level than most poor areas in China (Figure 2).

Since 2015, the Chinese government has exerted an unprecedented effort to reduce rural poverty and has initiated the strategy of targeted poverty alleviation (TPA). This strategy tackles rural poverty from income, infrastructure, public services, industrial development, and other more comprehensive aspects, thus contributing greatly to reducing the number of poor farmers [55]. Because of infrastructure, knowledge, technology, and other developmental elements sinking into rural areas, farmers today have more livelihood choices. However, at the end of 2016, ~30,000 farmers were still identified as poor by the government of Changchun. Moreover, farmers who have been lifted out of poverty are at risk of returning to this state, which has become a problem in China and several other countries [56]. Therefore, further research on farmers’ poverty, based on the local geographical context, is still urgently needed and has long-term significance not only for Changchun but also for similar areas.

3.2. Data

This study presents a case study of farmers’ households living in poverty that were defined by the government in TPA. Data were collected through structured interviews with poor farmers. The field survey was carried out between 2017 and 2018, and 1324 households were assessed (Figure 3). To gain comprehensive knowledge of the household situation concerning poverty and its underlying factors, questionnaires were divided into five parts. They started with the information on the heads of households, including name, gender, and education level. The second part was about household demographics, such as family size, age structure, physical health, and educational situation of school-age children. The third part included the condition of assets and living standards of cultivated land areas, housing, diet, and clothing. The fourth part focused on the income and income resources and these data were used to analyze income poverty and the livelihood strategy. The fifth part assessed policies and social security farmers received. The location of farmers was collected by GPS and then processed by AMAP to calculate distances to urban areas. Informal interviews with county officials and village cadres were also conducted to assess the situation of local development, the enterprises, processing plants, and cooperatives, etc. to analyze the nearby employment opportunities in villages and revise obtained household data. All data were processed by SPSS into numeric and categorical data.

3.3. Methods

The applied income poverty line matches the national rural poverty line of CNY2300 (constant price of 2010, USD ~1.6 per person per day). The FGT (Foster, Greer & Thorbecke) index was used to measure the poverty line, with alpha = 0, reflecting the poverty gap (PG). Because the income of specific households exceeded this poverty line, the formula was adjusted as follows to facilitate calculation. P represents the poverty line and Mi represents the annual per capita income of ith household. PG < 0 indicates that the household is experiencing income poverty, and the smaller the value, the stronger the income poverty.

$${\mathrm{PG}}_i=\frac{{\mathrm{M}}_{\mathrm{i}}-\mathrm{P}}{\mathrm{P}}$$

(1)

The indicator of multidimensional poverty index (MPI) refers to the MPI used by the UNDP and the Oxford Poverty and Human Development Initiative. This is combined with the TPA practice standards, including seven indicators of health, education, and living standards (

Table 1. Dimensions, indicators, and cutoffs of the multidimensional poverty index (MPI).

Dimensions	Indicators	Deemed as below the Poverty Line/Living in Poverty If:
Health	Physical condition	Any family member with a chronic illness, serious disease, or disability
	Medical insurance	Eligible household members do not participate in China’s rural medical system
Education	Years of schooling	Head of the household has not completed at least five years of schooling
	Child enrollment	Any school-aged child at the stage of compulsory education is not attending school
Living Standards	Housing	No long-term stable and safe housing or dwelling exhibits safety hazards
	Diet	Lack of staple foods, or protein available less than once per month
	Clothing	No seasonal clothing, shoes, and quilts for daily change or all have been donated

).

The child mortality rate in China is generally very low [57]. Few families in the assessed rural area have newborn children because of aging and emigration. Farmers’ self-reported health was used as part of the health dimension, which may arguably better reflect their actual health condition [58]. A further factor is farmers’ participation in the medical insurance system. China’s new rural cooperative medical system is a system of mutual assistance for farmers’ medical treatment that is organized and guided by the national government and requires voluntary participation of farmers. This system plays an important role in reducing medical expenses and ensuring farmers’ access to basic health services, thus guaranteeing their health.

All rural households that were surveyed in this study had electricity. Compared with electricity and floor standards, the comprehensive judgment of safety and stability for walls, roofs, and floors (in reference to the rural housing appraisal standard conducted by the Chinese government) was deemed a more accurate reflection of prevalent housing conditions, which is an important aspect in TPA [55,59]. Moreover, diet and clothing data used in the TPA were applied to reflect farmers’ living standards.

The Alkire and Forster (A&F) method was used to measure the number of deprived attributes for each household. To facilitate this calculation and analysis, Equation (2) was adjusted as below. Z_j represents the cutoff of indicator j, which was set as 1, meaning that the indicators were not deprived. Y_ij represents the value of household i’s j indicator. If indicator j is deprived, Y_ij is equal to 0; otherwise, it is equal to 1. The MPI value ranges from −7 to 0. If MPI = −7, all indicators are deprived; MPI = 0 implies that no deprivation occurs in any indicator.

$$\mathrm{MPI}={\displaystyle \sum }\frac{{\mathrm{Y}}_{\mathrm{ij}}-Z_{\mathrm{j}}}{Z_{\mathrm{j}}}$$

(2)

Logistic regression, cross-list analysis, statistical description, and qualitative analyses were used to explore the influencing mechanism of each factor on farmers’ poverty. Thirteen impact factors related to poverty were selected, which matches the approach taken by previous research and field surveys. The variance inflation factor (VIF) is the statistic for diagnosing collinearity in multiple regression. The larger the VIF, the more serious of collinearity between variables. Generally, the VIF should not exceed 10 [60]. In this study, the VIF of each factor was less than 3, and there was no multicollinearity problem between variables (

Table 2. Factors impacting poverty.

Impact Factors	Description	Min.	Max.	Mean	Tolerance	VIF
Amount of government subsidies provided	This includes agricultural subsidies, minimum living standard subsidy, disability subsidy, family planning subsidy, and veterans’ allowance	0.00	4.00	1.40	0.94	1.07
Cultivated land area (hm2)	Cultivated land area owned by the household	0.01	4.00	0.50	0.82	1.22
Livelihood strategy	1: none of the family members farming or working; 2: all members farming only; 3: any members both farming and working; 4: all members working only	1.00	4.00	2.06	0.71	1.40
Nearby employment opportunities	1: no employment opportunities, no enterprises, processing plants, and large cooperatives that can provide employment opportunities for villagers. 2: seasonal employment opportunities: farmers with large-scale farming or breeding work, or cooperatives requiring short-term and temporary workers. 3: long term employment opportunities: enterprises, processing plants, and cooperatives in the village that require long-term workers, which can provide relatively stable employment opportunities for villagers.	1.00	3.00	1.90	0.96	1.05
Gender of head of household	1: male; 2: female	1.00	2.00	1.20	0.97	1.03
Education level of head of household	1: primary school and below; 2: above primary school	1.00	2.00	1.23	0.97	1.03
Physical condition of family members	1: some members disabled and/or sick; 2: nobody disabled and/or sick	1.00	2.00	1.05	0.98	1.02
Family size	Number of family members	1.00	8.00	2.61	0.61	1.63
Proportion of people aged over 60 (%)	Number of people over 60/family size∗100%	0.00	100	48.71	0.77	1.30
Proportion of people with capacity to work (%)	Number of people with capacity to work/family size∗100%	0.00	100	28.72	0.97	1.03
Distance from center of Changchun (h)	The time distances from housing to centers under driving, calculated by AMAP, which is a widely used road navigation software in China and considers the road conditions in its calculation.	0.61	4.99	2.43	0.78	1.28
Distance from county center (h)		0.09	3.39	0.91	0.61	1.65
Distance from town center (h)		0.02	3.75	0.49	0.70	1.43
N	1324

).

4. Results

4.1. Income Poverty

4.1.1. Characteristics of Income Poverty

Most poor farmers in Changchun had incomes below or near the poverty line. To be specific, 41.09% of poor households presented income poverty. Households with income of less than half of the poverty line, i.e., serious income poverty, accounted for 14.12%. Among the 58.91% of households not showing income poverty, the incomes of 27.87% were 1.25 times lower than the poverty line, which implies that these households were very close to the poverty line. The PG mean was 0.376, and the income level of poor households exceeded the poverty line by an average of 37.60%. However, a large gap exists in the average income level of both groups, i.e., those indicating and not indicating income poverty. The PG mean of households with income poverty was −0.353. For the group without income poverty, the average PG value was 0.885, which substantially exceeded the poverty line (Figure 4).

4.1.2. Impact Factors of Income Poverty

Logistic regression results showed that in Changchun, governmental policies, non-agriculture employment, and demographic characteristics of families significantly impacted farmers’ income poverty (

Table 3. Results of logistic regression analysis on impact factors of income poverty.

Variables	Model 1: With Transfer Income			Model 2: Without Transfer Income
	B	Exp (B)	Sig.	B	Exp (B)	Sig.
Amount of government subsidies	−0.550	0.577	0.000	—		—
Cultivated land area	−0.102	0.903	0.000	−0.082	0.922	0.000
Livelihood strategy			0.000			0.000
Livelihood strategy (1)	2.497	12.147	0.000	3.859	47.431	0.000
Livelihood strategy (2)	1.560	4.757	0.000	1.932	6.903	0.000
Livelihood strategy (3)	−0.361	0.697	0.245	−0.236	0.790	0.350
Nearby employment opportunities			0.005	—		—
Nearby employment opportunities (1)	0.498	1.645	0.001	—		—
Nearby employment opportunities (2)	0.270	1.310	0.106	—		—
Physical condition of family members	0.900	2.460	0.011	—		—
Family size	0.501	1.651	0.000	0.349	1.417	0.000
Proportion of people with the capacity to work	—		—	−0.005	0.995	0.015
Distance from center of Changchun	0.005	1.651	0.000	—		—
Constant	−3.542	0.029	0.000	−0.754	0.470	0.009
R2	0.280			0.410
Prediction accuracy	58.9%			70.0%

B = logistic coefficient; Exp (B) = odds ratio; Sig. = significance; confidence level = 95%.

). The factors related to transfer income were complex and included the types, quantities, and standards of policies, as well as the number and financial situation of farmers’ adult children and other relatives. The externality of the transfer income may have weakened the explanatory power of the model. Therefore, the model was re-constructed after removing transfer income from the total income. The new model showed that although the number of major influencing factors decreased from seven to four, the R² and prediction accuracy improved significantly (Table 3).

Farmers who received more governmental subsidies were less likely to suffer from income poverty. Transfer income, in which governmental subsidies were dominant, and relatives’ financial support were minor parts, accounting for an average of 41.10% of the total income. Operating income, wages income, and property income accounted for 25.04%, 16.47%, and 17.39%, respectively. After removing the transfer income, the proportion of households with income poverty increased from 41.09% to 70.02%. The various subsidies that were directly paid to farmers from public finance expenditure were important for poverty reduction in farmers, which showed the direct effect the government had on farmers’ income.

Households that held more cultivated land were less likely to experience income poverty. However, more cultivated land contributed the least to reducing farmers’ income poverty in these two models, and the effect was reduced after removing the transfer income. This is strongly correlated with the falling corn prices and the single cropping structure which were greatly driven by macroeconomic regulation and control policies in trade and markets in the agriculture sector. In Changchun, farmland income was decreased overall by the price drop of corn. Furthermore, more than 90% of households with crop income grew corn only, which disabled them to offset this price impact. Therefore, cultivated land as a factor helping farmers to eliminate income poverty was not advantageous in this MGPA although its agriculture productivity is relativity high. More cultivated land areas implied more agriculture-related subsidies from the government. This may have resulted in the effective reduction of cultivated land in Model 2. Therefore, the effect of cultivated land on farmers’ income poverty was closely related to the policies, trade, and markets in agriculture sectors.

Factors related to non-agriculture employment, including diversified livelihood strategies and nearby employment opportunities, have contributed to the income poverty alleviation of farmers. Households that made their living neither by farming nor by working were always composed of the old and sick who had poor labor capacity and thus face a higher risk of income poverty. For households that made their living both by farming and working, when one income was affected, especially at a time when farmland income was affected, alternative incomes were used to compensate for the effect. Hence, these households were less likely to fall into income poverty. Households living in villages with employment opportunities, especially long-term employment opportunities, were less likely to experience income poverty. Employment opportunities in villages were generally basic skill and basic labor jobs farmers were familiar with or good at. Coupled with advantages of time flexibility and better accessibility in geographic space, nearby employment opportunities were always attractive and effective for farmers.

In family demographics, larger families, which had a higher dependency ratio because of more aging members and children, were more likely to show income poverty. The physical condition significantly influenced income poverty, but this was not the case after removing the transfer income. After the removal of transfer income, the number of households in poverty increased significantly both in groups with and without the disabled and/or sick. The increasing amount caused a larger proportion change in the latter group because of its small base. Therefore, the difference in poverty incidence between both groups had been narrowed, and the effect of physical condition was not significant as before. The proportion of people with the capacity to work followed the opposite trend. The data further showed that the gap in income poverty incidence between household groups with different proportion intervals of labor force widened, and the range increased from 6.35% to 14.67%. This indicates that after losing external support, the effect of the labor force on income poverty was emphasized for farmers.

4.2. Multidimensional Poverty

4.2.1. Characteristics of Multidimensional Poverty

The mean MPI was −1.97, and two attributes were found to be deprived in every household, on average. The households with MPI = −2 had the largest proportion and no households had MPI ≥ 6. The most severe problem was caused by the physical condition, with a deprivation rate of 96.00%. Among these households, 67.80% had sick members, 7.87% had disabled members, and 24.33% had both. The second most severe problem was caused by the years of schooling, with a deprivation rate of 76.74%, showing that most farmers had little education. Therefore, poor health and low education level were universal problems in rural Changchun. The third most severe problem was that of housing, where 15.42% of households had no long-term stable and safe housing or the dwelling exhibited safety hazards. Medical insurance (3.10%), diet (2.49%), clothing (2.34%), and child enrollment (1.21%) had significantly low deprivation rates (

Table 4. Proportions of MPI and deprivation rates of each indicator.

MPI		0	−1	−2	−3	−4	−5	−6	−7
Proportion (%)	Total	0.45%	21.22%	61.78%	14.20%	1.81%	0.53%	0.00%	0.00%
	Non-aging family	1.20%	26.62%	58.03%	12.71%	0.96%	0.48%	0.00%	0.00%
	Semi-aging family	0.00%	20.62%	64.52%	11.97%	2.22%	0.67%	0.00%	0.00%
	Aging family	0.22%	16.89%	62.50%	17.76%	2.19%	0.44%	0.00%	0.00%
Indicator		Physical condition	Medical insurance	Years of schooling	Child enrollment	Housing	Diet	Clothing	Mean MPI
Deprived rate (%)	Total	96.00%	3.10%	76.74%	1.21%	15.41%	2.49%	2.34%	−1.973
	Non-aging family	94.72%	3.36%	71.70%	1.68%	13.67%	1.20%	0.72%	−1.871
	Semi-aging family	96.23%	3.10%	75.39%	2.00%	14.86%	3.10%	3.10%	−1.978
	Aging family	96.93%	2.85%	82.68%	0.00%	17.54%	3.07%	3.07%	−2.061

Non-aging family: no households members were older than 60 years. Semi-aging family: part of the household members were older than 60 years. Aging family: all members were older than 60 years.

).

4.2.2. Impact Factors of Multidimensional Poverty

Poor health and low education level were strongly related to the backward production and living conditions of rural Changchun in the 20th century, when farmers were born and grew up. The original mode and low level of agricultural mechanization forced farmers to engage in high intensity and low safety manual labor, coupled with diets that were high in oil and salt in this area. This made osteoarthropathy and cardio-cerebrovascular diseases as well as disability by injury relatively common. This situation was also induced by the high demand for labor in agricultural production and to some extent, the neglect of education. The lagging medical and educational public services and poor infrastructure did not allow farmers to receive timely medical treatment and educational opportunities at an affordable cost.

Aging caused multidimensional poverty to deepen and to become more widespread. Semi-aging and aging families had significantly higher proportions when MPI ≤ −2, with more deprived indicators. Except for medical insurance and child enrollment, semi-aging and aging families had higher deprived rates in other indicators, especially aging families. Few non-aging families with better physical health and low medical needs took out medical insurance. In education, the nine-year compulsory education policy implemented by China in 1986 greatly promoted the popularization of school attendance for rural children. Today, families with children that had been removed from school were always those that were unable to attend school because of a physical condition, such as mental disabilities or deaf-mute.

In response to the rapid urbanization process, the rural aging problem became increasingly severe with the population outflowing from rural to urban areas [61]. The outflow of the younger and better-educated population aggravated farmers’ poverty and challenged rural poverty alleviation. In addition, another issue worthy of attention was that a number of farmers felt pressured by the increasing education expense because of the transfer of rural schools to townships or counties in response to the serious decline of the rural population. These schools always lie at a certain distance from remote villages, thus causing higher expenses for school buses, accommodation, and meals.

Income was found to be strongly correlated with multidimensional poverty. As shown by the results (Figure 5), the smaller the number of deprived attributes in multidimensional poverty, the smaller the proportion of households trapped in income poverty, and the lower the degree of income poverty. This correlation becomes more significant with increasing living standards. Households with deprived living standards were mainly those experiencing income poverty, some of which potentially exceeded the poverty line while remaining very close to it. The PG means of households that were deprived in relation to housing, diet, and clothing were 0.170, −0.299, and −0.374, respectively, which were clearly lower than the average. The cost of improving housing conditions is relatively high and difficult to afford for certain households whose income is just slightly above the poverty line. This may be the main reason why the group with deprived housing was relatively large.

5. Discussion

In areas with limited natural resources and poor ecological environments, where farmers always face challenges of poor quality and quantity of farmland, and may even experience hunger, increasing agriculture productivity was always an effective means to alleviate poverty [11,15,40]. While agriculture is important, agriculture alone does not reduce poverty [24,26]. This study shows that in MGPAs with a better natural environment and relativity high agriculture productivity, governmental power and economic factors play more important roles in farmers’ income. Better agricultural production conditions do not emphasize the role of cultivated land in poverty reduction, and a higher economic level does not generate comparable income levels for rural residents in Changchun. This means that farmers can hardly share the economic growth and obtain compatible benefits in agriculture sectors because of the uneven distribution of economic benefits among sectors [25,49]. However, farmers who are faced with high market risk bear the cost of trade and competition by reducing agricultural income. The reason is that global trade liberalization and market competition in the agricultural sector greatly impacted farmers’ income at the micro-level through agricultural prices [44,46]. Therefore, agriculture and farmers’ income are associated with low benefits but high risks. Combining trade reforms of different countries and macroeconomic regulation of different sectors is necessary [22]. Some measures are the potential to reduce farmers’ poverty in MGPA. These include tilting the benefits of economic growth to farmers by increasing/stabilizing the prices of agricultural products, diversifying planting structure, and transferring the market cost to subjects of enterprises and factories that with higher risk resistance ability in segmented markets.

Aging showed different correlations with poverty in farmers. In contrast to the USA, where aging farmers still hold large farms and achieve agricultural modernization, or the EU, where farmers can receive considerable subsidies that sometimes make them face a lower poverty risk [31,51], in China and other developing countries with small scale agriculture operation and relativity low level of agricultural modernization, aging generally implies associated income decrease as well as wider and deeper poverty levels [44]. Moreover, aging results in the deficiency and weakening of rural construction bodies, thus restricting rural development [62]. Farmers operate 36.90% of all land area on Earth as cultivated land and are responsible for the food security of humanity [63], especially those in areas such as MGPAs. Therefore, aging in such areas will aggravate problems of labor force shortage in agriculture and food security. Who will become the farmers to operate agricultural production and whether food security can be guaranteed are key challenges [50]. Aging is an inevitable trend. Improving the human capital of aging farmers by strengthening their advantages and weakening their disadvantages is a potential direction. With the extension of the human lifespan, certain farmers over 60 years of age do not lose their labor ability completely, and they are experienced and familiar with local agriculture. Enhancing health education and medical services for farmers to improve their physical quality, combed with improving agricultural technology, mechanization, and vocational training will help reduce the poverty risk and alleviate labor shortage caused by aging, especially for the new generation of older farmers. The education resources in local universities, especially those related to agriculture should be fully utilized by encouraging teachers and students to study on farmland with farmers. This can achieve improvement and updating of farming skills and knowledge of farmers.

As a disadvantaged group, poor farmers face inherent obstacles of aging, poor physical health, and little education. This makes it difficult for them to utilize favorable conditions of urban driving forces that enable access to employment opportunities, markets, and services [28,52,64,65], ultimately causing them to become a poor group in a wealthy area [1,20]. Many transfer payments from public finance expenditure play an important role in reducing the poverty level of those farmers by providing various direct subsidies. However, in the long run, subsidy-based measures for poverty alleviation impose a heavy burden on governmental finances and become unsustainable, especially in areas with large populations. Such measures consume a large share of the government budget and distract from growth-enhancing investments [22], even making farmers mentally ”relying on” subsidies [56]. Thus, the quality of public spending—the efficiency of resource use—is often an even more important issue to address than its level [22]. In MGPAs that have the advantages of metropolitan areas, public finance expenditure can develop effective poverty reduction measures in more fields. These include improving land quality as well as supporting the development of agriculture-related enterprises and factories to promote the integrated development of agriculture and other industries. These measures will help farmers, especially those with poor labor ability, get higher asset returns and lower poverty risk in these processes.

The concept of unequal spatial opportunity structures offers a potential avenue for a better understanding of the drivers of social disadvantage, and poverty has already been confirmed in different regional and national contexts [3]. Similar to the study by Bernard on the Czech Republic, the present study also confirms that the residential disadvantage was relevant to farmers’ poverty from a micro-regional perspective [52,53]. Poor farmers living in villages that offer employment opportunities were less likely to experience income poverty. Different production and living conditions and socio-economic developments in different regions have created different livelihood environments for farmers, thus affecting their income levels [5,8,11]. Therefore, except as an objective spatial carrier to maintain livelihoods, the location of farmers is also a kind of spatial capital that entails differences in development opportunities that farmers can utilize. Improving the capital value of rural space is a potential measure to improve farmers’ livelihoods and reduce farmers’ poverty. This includes optimizing rural production and living conditions by optimizing infrastructures and facilities to promote the attractiveness of rural areas for enterprises and factories [17,33] as well as developing rural tourism to enhance the added value of rural resources [47,66]. Integrating farmers themselves as the main body in those process enterprises is important. However, spatial opportunity imposes certain requirements on human capital [64,65]. Under a low human capital level, spatial opportunity is not effective for poverty reduction.

6. Conclusions

This study is based on a field survey of poor farmers in one of China’s MGPAs, i.e., Changchun, which is an area type that was previously overlooked by farmers’ poverty studies. An analysis of income poverty and multidimensional poverty showed that the main issues associated with farmers’ poverty in this area include Governmental policies and the macroeconomic situation in the agricultural sector, non-agricultural employment, aging, cultivated land, and family size.

In the MGPA of Changchun, the multidimensional poverty of poor farmers is more widespread than income poverty. The income of poor farmers greatly relied on government subsidies greatly, which have reduced income poverty by more than 30%. The potential of favorable agriculture development in poverty reduction was impacted by governmental policies and the economic development situation behind these policies through market prices and subsidies in agricultural sectors. Non-agriculture and mixed livelihood can reduce income poverty, and such livelihood opportunities are more effective within villages than in urban areas because of the obstacles imposed on poor farmers by their reduced labor capacity. In MGPAs, farmers can be grain self-sufficient, which reduces the deprivation of living standards and improves income, which also reduces the deprivation of living standards. Aging is a key cause that further deepened farmers’ poverty levels and made farmers’ poverty more widespread, especially for farmers with low education levels and poor physical conditions. These are the most serious and prevalent problems poor farmers face.

By examining MGPAs, this paper expands the regional types that rural poverty research has tended to overlook. This expansion is conducive to a more comprehensive understanding of poverty and its diversity with changing geographical context. This is also helpful to proceed with poverty alleviation policies in a more targeted way and according to regional differences, highlighting the spatial thinking in poverty governance. This research assesses poor rural households, which provides a basis for the further exploration and analysis of the characteristics and impact factors of poverty from the perspective of the family unit. This perspective is more micro-focused and thus, closer to poor farmers themselves.

Farmers’ poverty is a process that dynamically changes with population structure and developments of society and economy. This dynamic process could not be elicited through the cross-sectional household survey data used in this paper. The effect of relevant impact factors on farmers’ poverty, i.e., policies, is sustained and possibly lagging. Although farmers’ poverty has been reduced considerably over the last several years because of the notable effort of TPA, especially regarding medical insurance, child enrollment, housing, diet, and clothing, all of which can be improved in a short time, farmers’ lives have faced and continue to face many objective difficulties and risks. Examples are demographic structural problems of serious disease and disability, low educational level, aging, and socio-economic problems of unstable markets, slow economic growth, uneven benefit distribution among sectors, and instabilities associated with the market and the economy. The increasingly complex situation in global trade, the epidemic of Covid-19, war, etc., are exacerbating these problems further. Farmers face a more changeable environment and even higher poverty risks. Therefore, it is necessary to conduct more targeted research in the future. In the research object, taking both poor farmers and non-poor farmers as research objects will help to form a more comprehensive cognition of farmers’ poverty. Regarding methods, follow-up field surveys and observations must be conducted, and sequential research must be done through years of data accumulation to understand the dynamics of farmers’ poverty and their underlying mechanism. At the research scale, a macro-scale of farmers’ poverty can be studied by integrating field survey data, census data, socio-economic data, spatial and remote sensing data, and by mapping farmers’ poverty at different spatial scales. How socio-economic and land-use changes affect farmers’ poverty are important and valuable future research directions.