Why Do Farmers Support Stable Land Ownership? Marketization with Chinese Characteristics

Abstract

Recent debates regarding marketization have focused on the relationship between the state and the market, while the grassroots and their everyday experiences have arguably been understudied. In this paper, we study marketization with the example of land marketization in China. Out of concern for the grassroots’ perspective, we investigate Chinese farmers’ perceptions with regard to stable land ownership of farmland, which are essential for land marketization in the backdrop of intensive land use conversion in China’s urban periphery. Approximately 1200 farmer households were interviewed around 12 cities in mainland China. An ordered probit regression analysis of the survey results reveals a series of factors that explain the individual farmers’ preference for stable land ownership. Among others, the decreasing size of farmer household and rural women’s insecure property rights in farmland are identified as two grassroots-based characteristics underpinning China’s ongoing transition to a more market-based farmland use institution. An important theoretical implication of our research is that the mainstream literature perhaps over-attributes China’s marketization to the state and the market, while under-evaluating the spontaneous support from bottom-up.

1. Introduction

In the broad international development context, the last few decades have witnessed two different, albeit related, types of critical reflections on the global spread of marketization. The first kind of critique concentrates on the undesirable outcomes of market-oriented policies, mainly in terms of the intensification of socioeconomic inequalities and environmental degradation [1,2,3]. By comparison, the second group of scholars employ typically a Habermasian ideologiekritik approach to marketization, challenging its epistemological obsession with the market-state nexus, alongside its oversight of the grassroots [4,5,6].

The market-oriented reform in China offers a case to examine marketization against the background of economic globalization. The marketization of land, especially that of farmland, is one of the fundamental reforms in China’s marketization reform. Farmland conversion in China reflects a variety of fundamental socioeconomic transitions in contemporary China and relates closely to the debate about whether the country has seen a sustained process of marketization since the reforms of the late 1970s [5,6,7]. On one hand, the reform of Household Responsibility System (HRS), characterized by the reallocation of farmland use rights to individual families, may be seen as a typical market-based “structural adjustment” [5,6,7]. On the other hand, the Land Management Law restricts almost all kinds of non-agricultural use of farmland to the eminent domain (with the exceptions of building town and village enterprises and village infrastructure), entitling local governments to convert farmland through land requisition [8].

In the literature on farmland conversion in China, there is also an intellectual divide regarding the global spread of marketization. Most existing research, arguably, seems to focus on the role of the state and the market in the marketization of land, often employing extensive analyses of the structure of farmland usufructs under the HRS and the various social and environmental issues related to the exercise of land requisition [9,10,11]. In contrast, a relatively small number of publications attempt to move beyond the state-market dichotomy to look at farmland conversion from a grassroots perspective that emphasizes the farmers’ everyday cultural experiences [12,13,14].

In this paper, we continue to interrogate the dichotomist market-state discourse about farmland conversion in China, albeit through a statistical analysis in lieu of the qualitative approach that is more conventional in this kind of research. Based on a large set of survey data collected between 2008 and 2009 with 1209 farmer households across 12 localities in mainland China, our regression analysis suggests that rural grassroots opinions depend neither purely on the state nor the market. While a majority of the respondents appeared to prefer stable land ownership of farmland, their preference was driven by very specific localized factors, such as the shrinking size of farmer households due to local demographic transitions and female farmers’ insecurity in terms of their post-marriage property rights in farmland. Given these findings, we argue that the marketization of China is not only attributable to the general state-market relationship, but also to some highly place-based local Chinese characteristics.

The remainder of our paper is organized as follows. The next two sections review the academic debate about China’s marketization and its relation to the academic research on farmland conversion and stable land ownership in China’s urban periphery. Following that is a presentation of the survey design and regression method employed in this study, with the model results reported and analyzed afterwards. Before the conclusion, the policy and academic implications of this research are discussed in detail.

2. Marketization, Developmental State, and Beyond

Recent reflections on marketization in the world primarily focus on the relation between the state and the market and, most inherently, epistemological structuralism [15,16]. These ideologiekritiks follow a quite different line of thinking from that of previous studies, which are primarily concerned with the substantive consequences of marketization policies, such as the intensification of socioeconomic inequalities and environmental degradations [17,18,19,20]. The fact that major international organizations nowadays have increased their funding support for Third-World-based nongovernmental organizations more or less reflects a determined departure from the old structuralist state-market dualism [21]. The role of the individual in marketization also needs to be emphasized. These are three main perspectives of marketization reform in different countries (see

Table 1. Different perspectives on international development.

	Marketization	Developmental State	Grassroots’ Agency
Philosophical foundation	structuralism	structuralism	agency & structure
Epistemological focus	How does the market influence the structure of civil society?	How does the state proactively regulate the market and influence the structure of the economy and civil society?	How does civil society interact with various constraining structures (including but not limited to the state and the market)?
Typical policy recommendations	free trade	regulation; trade protection;	place-based governance; local economic development
Major research subject	the market structures	the state institutions	the grassroots’ perceptions and actions
Typical authors	Oi, J. [15]; Tong, Y. [16]	Cowling, K.; Tomlinson, P.R. [19]; Narins, T. [20]	Sulistyani, W [21]
Limitations	overemphasize the effect of the “invisible hand”(market) and ignore the role of state	believe that the state should regulate the market development	grassroots’ forces cannot take effect without support from the state.

).

As a large developing country, China’s market-oriented reform offers an example for the study of marketization against the background of globalization. The reform, which started in the late 1970s, was an institutional reform that aimed to transition from a planned economy to a market economy and establish a market economy system regulated by the state. The process of China’s marketization has aroused much attention. Some studies are dedicated to measuring and accessing China’s marketization levels with various aspects and indexes, including the proportion of economic resources allocated by the market, development of non-state-owned economy, government size, and so on [22,23,24]. These studies focus on the role of the developmental state and claim that state and government play the most important roles in the marketization process. The regulation of the state on the market has guided the market in the right direction, and the protection of trade order ensures a healthy market environment. Under China’s communal land ownership system, land marketization depends more on state and government. Other studies pay attention to the influence of marketization on economic or social conditions, for example, on the income gap [25], economic growth [26,27], the implementation of social policy [28], social trust [29], etc. These market-centered theses demonstrate that marketization is the driving force of China’s social development after the reform and opening-up. The allowance of free trade unleashed the dynamism of China’s economic market and promoted social change. Allowing the free trade of land was the most important breakthrough. Land marketization plays a crucial role in the process of China’s marketization. The major debate between these two perspectives is whether the state or the market dominates the process of land marketization in China. Further studies focus on the interaction between marketization and other factors, for instance urbanization, corruption, regional heterogeneity, decentralization, globalization, etc. [30,31,32,33,34,35]. These studies ignore the interaction between the state and market. Subsequently, some scholars focus on their interaction. Research on China’s marketization also sees an overwhelming concentration on the role and interaction of the market and the state. In comparison, an attention to the grassroots is rare. While there are some studies that have extensively explored grassroots experiences and behaviors in marketization [23], the research on the opinion and perception of farmers about marketization is still rare, which may have influence on marketization.

The marketization of land, especially that of farmland, was one of the fundamental reforms in China’s marketization reform. Land reform in China was initiated by the reform of the HRS in 1978, which reallocated the use-right of farmland to individual households and allowed the transaction of farmland. Then, a series of reforms, including the reform of farmland ownership confirmation and the “separation of three rights”, guaranteed farmland to be freely transferable within agricultural use, while the transfer of farmland into nonagricultural use could only be achieved by land expropriation and land transfer of government. China’s land reform has aroused much attention. Some scholars pay attention to the process and implication of land reform [36,37,38]; others focus on the impact of land reform on individual behaviors, for example the relation between land tenure reforms and land conservation investment [39]; others dedicate their studies to the effects of land reform on a macro level, for instance, on the scale of land use and rural development [40,41].

In the area of land reform, there is also an overwhelming concentration on the role and interaction of the market and the state. Land marketization reform in China, which originated with the reform of the household responsibility system in the late 1970s, was considered a state-guided reform from the very beginning [37]. The continued marketization of farmland is largely dependent on the reform of the state [38]. In the specific practice of land marketization, for example in farmland conversion, the state has a great impact on the degree of marketization [39,40]. For these reasons, land reform in China is portrayed by many scholars as a process of marketization that relies on the state’s role in the market and is driven largely by government policies [42]. Despite the primary role that the state plays in land marketization, the market has also seen its growing importance in the process of land reform. For one thing, market capitalism is constantly promoted in the process of land marketization. For another, regulated as they are by the land market, land transactions are becoming increasingly common and normalized in rural areas, especially in the way of land lease.

3. Farmland Conversion and Stable Land Ownership in China

In China, marketization is one of the most conspicuous reforms since the late 1970s. It sounds to establish a market economy system regulated by the state, and to form a unified market operation mechanism and market system. In the 40 years since the reform and opening-up, one of the most important elements of China’s economic and social development is the promotion of marketization with property rights reform. In particular, reform of rural land property rights is an important foundation for the marketization of land, which indicates the market allocation of land factors. With the ongoing reform, land ownership has become more stable and the transfer of rural land within agricultural use has become increasingly common.

Stable land ownership is essential for land transfer. In China, the stability of land ownership is reflected in many aspects, for example in land certificate, land adjustment, and land finance [43,44]. All of these are important components of land reform. It has been confirmed that stable land ownership helps to improve the scale and efficiency of land transfer by reducing transaction costs and improving land value [45,46]. With the continuous land reform over the years, the transactions of farmland within agricultural use has become more common in rural areas. However, the conversion of farmland into non-agricultural use is still strictly restricted by the state and can only be achieved by land requisition. According to Land Management Law, land requisition is the only legitimate way for farmland to be converted into non-agricultural use (with the exceptions for building town and village enterprises and village infrastructure). By land requisition, collectively owned land is transformed into state-owned land; then, once approval of land conversion has been attained, the original farmland can be used for non-agricultural use. After land requisition, farmers lose their land ownership permanently, which comprises a threat to the stable ownership of farmland.

In this sense, farmers’ preference for stable land ownership indicates their longing for a smoother transaction of land use rights, and also their reluctance surrounding land requisition (instead of land conversion by market). Taken together, this reflects farmers’ preference for land marketization.

There has been a large amount of research conducted on the marketization of China. Although recent decades have witnessed an increasing number of sophisticated and well-reasoned academic critiques regarding China’s reform from both the angles of the state and the market [21,22,47], grassroots-based discourses remain relatively nascent.

This research gap is especially evident in studies about the marketization of land. Most of the literature seems to concentrate on the continued marketization of farmland following the household responsibility system reform in the late 1970s [48,49]. Another main research stream focuses on the role of the state in the marketization of land, mainly in the way of farmland conversion [50,51]. By comparison, discourses about the marketization of land from a farmer’s perspective are on the minority side and, in terms of their research methods, rely overwhelmingly on ethnographic fieldwork [25,26,27]. Quantitative scholarship based on systematically-collected empirical data is a rarity. To fill this academic gap, we raise two essential research questions in our paper:

(a)

What are the individual farmers’ perceptions about the farmland usufruct under the HRS (Household Responsibility System)?

(b)

Do farmers prefer more stable land ownership vis-à-vis the existing land requisition system to convert farmland into urban uses?

To address both questions systematically, we employ a rigorous social survey and statistical analysis approach to assess Chinese farmers’ viewpoints on stable land ownership against the background of rapid farmland conversion in China’s urban periphery.

4. Research Methods

4.1. The Sampling Framework

Between December 2008 and August 2009, we conducted a multistage geographical cluster sampling and interviewed 1209 farmers from different households around 12 cities (C1 to C12 in Figure 1) across east, north, south, and west China (see Figure 2 and

Table 2. Sample distribution and interviewees’ demographic information.

City	Sample Size	Male	Farmer	Ethnic	Party Member	Local Officials	Veteran
Yueqin (C1)	81	72.84%	96.3	0.00%	16.05%	4.94%	12.35%
Ninbo (C2)	102	78.43%	63.73	0.00%	40.20%	6.86%	12.75%
Jiangyin (C3)	101	80.2%	79.21	0.00%	8.91%	3.96%	8.91%
Sanhe (C4)	108	62.04%	87.04	0.00%	31.48%	7.41%	13.89%
Weifang (C5)	109	73.39%	84.4	0.00%	19.27%	4.59%	7.34%
Jinan (C6)	102	59.80%	95.1	0.00%	24.51%	9.80%	7.84%
Guangzhou (C7)	105	77.14%	42.86	0.00%	37.14%	3.85%	13.33%
Zhongshan (C8)	86	74.42%	5.81	0.00%	31.40%	5.81%	13.95%
Dongguan (C9)	90	76.67%	87.78	0.00%	16.67%	5.56%	11.11%
Chongqin (C10)	106	46.23%	85.85	0.00%	11.32%	6.60%	1.89%
Nanchong (C11)	108	66.67%	47.22	0.00%	11.11%	4.67%	2.78%
Chengdu (C12)	111	74.77%	73.87	0.9%	28.83%	7.21%	8.11%
Overall	1209	69.98%	71.05	0.08%	23.16%	5.97%	9.35%

). The survey was conducted in four major regions (Yangtze River Delta, Pearl River Delta, Yellow Sea and Bohai Sea, Chengdu Chongqing), which are located in east, south, north, and west China, respectively. According to the principle of stratified sampling, all cities of each region are classified into large, medium, and small ones, and one city is randomly selected from each category to generate 12 cities. In each city, two townships were randomly selected. Then from each township, we randomly chose five villages in the urban periphery, as the degree of marketization is higher in this place and the farmers have a stronger conception on marketization. Finally, about ten households were randomly selected in each village for one-to-one questionnaire interviews. Table 2 illustrates the sample distribution across the 12 cities, alongside the respondents’ basic demographic information.

4.2. The Questionnaire

Our survey questionnaire was structured with respect to three key interview questions (see

Table 3. Key survey questions regarding stable land ownership of farmland.

Question 1	Do you support more stable ownership of the collective farmland under the Household Responsibility System (HRS)?
Question 2	If the ownership of farmland is more stable, it would be more difficult to adjust each household’s access to farmland based on the changing size of household under HRS. Do you think that is acceptable and appropriate?
Question 3	If the ownership of farmland is more stable, it can be transacted more freely in the market. Do you think that is acceptable and appropriate?
Measurement	−1 = No; 0 = Uncertain; 1 = Yes

), which were deployed collectively to measure the dependent variable, i.e., whether a respondent agrees with stable land ownership or not. Farmers’ preference for stable land ownership is an indicator of their opinion on farmland marketization, as stable land ownership is necessary for the transaction of land use rights. Question 1 is an overarching, though abstract, query regarding an individual farmer’s overall opinion about stable land ownership. In contrast, questions 2 and 3 draw on more specific economic interests related to the issue. The purpose of these two specific questions is to make sure farmers have good knowledge of the effects of stable land ownership and what it means to marketization. Question 2 reads as follows: “If the ownership of farmland is more stable, it would be more difficult to adjust each household’s access to farmland based on the changing size of household under HRS. Do you think that is acceptable and appropriate?” In China, farmland was distributed equally among members of the collective and used to be adjusted according to demographic changes. Stable land ownership makes it more difficult to adjust land ownership according to family population growth, which means a loss for those households with expected population growth. This is the main reason that hinders farmers from agreeing to stable land ownership. Question 3 reads: “If the ownership of farmland is more stable, it can be transacted more freely in the market. Do you think that is acceptable and appropriate?” The free transaction of land is at the core of land marketization, and it has been confirmed that stable land ownership has a positive impact on rural land circulation by reducing transaction costs and improving land value [52,53]. If a respondent answered “yes” to questions 2 and 3 after saying “no” or “uncertain” to question 1, we would ask the person to revisit question 1. We would do the same if a farmer answered “no” to both questions 2 and 3 after saying “yes” or “uncertain” to question 1. Otherwise, we simply filed the original answer to question 1 as the result.

We also asked questions about each individual farmer’s perception about land requisition because we assumed it to be an important factor in explaining whether a farmer would prefer stable land ownership or not. As well as this information,

Table 4. Supplementary questions in the survey.

Variable	Question and Measurement
Land Requisition (V1)	Has a land requisition taken place during 2004–2008? 1 = Yes; 0 = No
Dissatisfied with Requisition (V2)	Are you dissatisfied with the compensation for farmland requisition? 1 = Yes; 0 = Uncertain
Satisfied with Requisition (V3)	Are you satisfied with the compensation for farmland requisition? 1 = Yes; 0 = Uncertain
Land Reallocation (V4)	Have you ever experienced farmland reallocation under HRS? 1 = Yes; 0 = No
Age (V5)	Age in 2008
Gender (V6)	1 = Male; 0 = Female
Education (V7)	Years of full-time education
Household Size (V8)	Number of people per household in 2008
Change in Household Size (V9)	Has the size your household changed since the last round of farmland reallocation under HRS? 1 = Increased; 0 = Otherwise
Off-farm Income (V10)	Off-farm income in 2008/total household income in 2007
Per Capita Farmland in mu (V11)	Amount of arable land per capita within the household
Per Capita Income in RMB (V12)	Per capita income within the household
Party Member (V13)	Are there any members of the Chinese Community Party within the household? 1 = Yes; 0 = No
Local Official (V14)	Are there any local officials within the household? 1 = Yes; 0 = No
Veteran (V15)	Are there any veterans within the household? 1 = Yes; 0 = No
Working elsewhere (V16)	Are there any household members working elsewhere? 1 = Yes; 0 = No

lists all the explanatory variables we managed to collect data on through the survey.

4.3. The Regression

We applied an ordered probit regression method to analyze the survey results. Following Greene, we tested the respective probabilities of three responses with regard to an individual farmer’s unobserved latent preference for stable land ownership.

$$\mathrm{Prob}(y_i=-1|x)=\prod ({\alpha }_1-x\beta )$$

(1)

$$\mathrm{Prob}(y_i=0|x)=\prod ({\alpha }_2-x\beta )-\prod ({\alpha }_1-x\beta )$$

(2)

$$\mathrm{Prob}(y_i=1|x)=1-\prod ({\alpha }_2-x\beta )$$

(3)

Equations (1)–(3) estimate, respectively, the probability for an unobserved farmer, i, to disagree, be uncertain, or agree with stable land ownership, given a matrix of empirical observations, x, and a vector of assessable coefficients, β. ∏ denotes the cumulative probability function corresponding to the standard normal distribution, while α₁ and α₂ stand for two threshold values or cut-off points on ∏ that are subject to estimation.

We then specified two linear regression models to test the three probabilities mentioned above. In the first model, we analyzed all of the 1209 sample observations based on all of the explanatory variables included in Table 4, supplemented by 12 dummy control variables (i.e., C1 to C12), each corresponding to one of the 12 cities in Table 2. We followed the same approach in testing the second model, except that this time we only studied those who had experienced at least one farmland requisition between 2004 and 2008.

5. Empirical Findings

5.1. Descriptive Statistics

Table 5. Descriptive statistics about farmers’ preference for stable land ownership.

City	Cases	Do You Support Stable Land Ownership?
		Agree	Percent	Disagree	Percent	Uncertain	Percent
Yueqin (C1)	81	34	41.98%	20	24.69%	27	33.33%
Ninbo (C2)	102	56	54.90%	38	37.25%	8	7.84%
Jiangyin (C3)	101	60	59.41%	40	39.60%	1	0.99%
Sanhe (C4)	108	55	50.93%	41	37.96%	12	11.11%
Weifang (C5)	109	39	35.78%	65	59.63%	5	4.59%
Jinan (C6)	102	41	40.20%	57	55.88%	4	3.92%
Guangzhou (C7)	105	43	40.95%	56	53.33%	6	5.71%
Zhongshan (C8)	86	50	58.14%	32	37.21%	4	4.65%
Dongguan (C9)	90	42	46.67%	46	51.11%	2	2.22%
Chongqin (C10)	106	45	42.45%	46	43.40%	15	14.15%
Nanchong (C11)	108	50	46.30%	52	48.15%	6	5.56%
Chengdu (C12)	111	59	53.15%	49	44.14%	3	2.70%
Overall	1209	574	47.48%	542	44.83%	93	7.69%

reports the descriptive statistics regarding the dependent variable, i.e., whether an individual farmer supports stable land ownership or not. It is conceivable that a higher percentage (47.48% vs. 44.83%) of the farmers in our sample agreed to more stable land ownership of the farmland.

Table 6. Independent variables’ descriptive statistics (sample size = 1209).

Variable	Cases	Mean	Sth	Min	Max
Land Requisition (V1)	1209	0.680728	0.466388	0	1
Dissatisfied with Requisition (V2)	823	0.307412	0.461702	0	1
Satisfied with Requisition (V3)	823	0.244228	0.42989	0	1
Land Reallocation (V4)	1209	0.315964	0.745462	0	8
Age (V5)	1209	50.08768	11.60222	18	87
Gender (V6)	1209	0.699752	0.458556	0	1
Education (V7)	1209	7.732837	3.273755	0	19
Household Size (V8)	1209	3.8933	1.423326	1	8
Change in Household Size (V9)	1209	0.047974	0.213799	0	1
Off-farm Income (V10)	1209	0.038958	0.195833	−3.42466	1
Per Capita Farmland in mu (V11)	1209	0.217085	0.512763	0	5
Per Capita Income in RMB (V12)	1209	15766.83	20578.99	−14350	241666.7
Party Member (V13)	1209	0.331679	0.471011	0	1
Local Official (V14)	1209	0.080232	0.271764	0	1
Veteran (V15)	1209	0.179487	0.383919	0	1
Working elsewhere (V16)	1209	0.630273	0.482931	0	1

summarizes the descriptive statistics about the main independent variables, though it excludes the geographic control variables. As only 823 of the 1209 respondents had experienced farmland requisition between 2004 and 2008, the sample size for V2 and V3 shows only as 823 in Table 6.

Table 7. Local breakdown of respondents who experienced land requisition.

City	Cases	Cases V1 = 1	Proportion of Cases with V1 = 1
Yueqin(C1)	81	61	75.31%
Ninbo(C2)	102	87	85.29%
Jiangyin(C3)	101	80	79.21%
Sanhe(C4)	108	54	50.00%
Weifang(C5)	109	71	65.14%
Jinan(C6)	102	98	96.08%
Guangzhou(C7)	105	38	36.19%
Zhongshan(C8)	86	20	23.26%
Dongguan(C9)	90	35	38.89%
Chongqin(C10)	106	74	69.81%
Nanchong(C11)	108	95	87.96%
Chengdu(C12)	111	110	99.10%
Overall	1209	823	68.07%

illustrates the local proportion of respondents who had experienced farmland requisition (i.e., those who report V1 = 1) across the 12 sample cities. It is conceivable that those cities that show a higher percentage of farmer respondents who had experienced land requisition also tend to display a higher proportion of farmers who support stable land ownership, for example, Ninbo (C2) and Jiangyin (C3) in east China.

This prompted us to focus on the 823 respondents who had experienced at least one farmland requisition between 2004 and 2008.

Table 8. Independent variables’ descriptive statistics (sample size = 823).

Variable	Cases	Mean	Std	Min	Max
Land Requisition (V1)	823	0.307412	0.461702	0	1
Dissatisfied with Requisition (V2)	823	0.244228	0.42989	0	1
Satisfied with Requisition (V3)	823	0.368165	0.81906	0	8
Land Reallocation (V4)	823	49.33779	11.90729	18	87
Age (V5)	823	0.712029	0.453093	0	1
Gender (V6)	823	7.880316	3.35935	0	19
Education (V7)	823	3.73147	1.422642	1	8
Household Size (V8)	823	0.053463	0.225092	0	1
Change in Household Size (V9)	823	0.027003	0.171028	−3.42466	1
Off-farm Income (V10)	823	0.112067	0.296879	0	3.19
Per Capita Farmland in mu (V11)	823	15,839.72	21,421.61	−1136	241,666.7
Per Capita Income in RMB (V12)	823	0.339004	0.473659	0	1
Party Member (V13)	823	0.072904	0.260137	0	1
Local Official (V14)	823	0.18955	0.392184	0	1
Veteran (V15)	823	0.643985	0.479111	0	1

is the counterpart of Table 6 after we focused on the 823 farmers who had had their farmland expropriated.

Table 9. Local breakdown of farmers’ satisfaction with land requisition.

City	Obs	Satisfaction with Compensation for Land Requisition
		Satisfied (V3 = 1)	Percent	Dissatisfied (V2 = 1)	Percent	Unsure (V2 = 0 = V3)	Percent
Yueqin(C1)	61	11	18.03%	40	65.57%	10	16.39%
Ninbo(C2)	87	29	33.33%	40	45.98%	18	20.69%
Jiangyin(C3)	80	39	48.75%	19	23.75%	22	27.50%
Sanhe(C4)	54	31	57.41%	15	27.78%	8	14.81%
Weifang(C5)	71	56	78.87%	2	2.82%	13	18.31%
Jinan(C6)	98	57	58.16%	26	26.53%	15	15.31%
Guangzhou(C7)	38	8	21.05%	12	31.58%	18	47.37%
Zhongshan(C8)	20	2	10.00%	6	30.00%	12	60.00%
Dongguan(C9)	35	4	11.43%	4	11.43%	27	77.14%
Chongqin(C10)	74	23	31.08%	29	39.19%	22	29.73%
Nanchong(C11)	95	51	53.68%	30	31.58%	14	14.74%
Chengdu(C12)	110	58	52.73%	30	27.27%	22	20.00%
Overall	823	369	44.84%	253	30.74%	201	24.42%

then shows a local breakdown of the farmers’ satisfaction level about the compensation they had received for the latest land requisitions.

5.2. Regression Results

In this section, we chose ordered probit regression as our study method. Ordered probit regression is a special method of ordinal regression that is suitable for data analysis where the dependent variable is an ordered categorical variable. The dependent variable of our study - farmers’ preference for stable land ownership - is a categorical variable with three categories (agree, disagree, uncertain). The three categories have a certain order, which meets the requirements to use this method.

The results of two ordered probit regressions are presented in

Table 10. Results of two ordered probit regressions.

Explanatory Variables	Regression Model 1	Regression Model 2
Land Requisition (V1)	0.472 ***
	(0.0966)
Dissatisfied with Requisition(V2)		0.249 **
		(0.108)
Satisfied with Requisition (V3)		0.0224
		(0.117)
Reallocation (V4)	0.0217	−0.0244
	(0.0538)	(0.0595)
Age (V5)	−0.000538	−0.00391
	(0.00381)	(0.00474)
Gender (V6)	−0.263 ***	−0.248 **
	(0.0803)	(0.0987)
Education (V7)	0.00287	0.00494
	(0.0124)	(0.0148)
Household Size (V8)	0.0222	0.0238
	(0.0286)	(0.0352)
Change in Household Size (V9)	−0.539 ***	−0.615 ***
	(0.175)	(0.195)
Off-Farm Income (V10)	0.165	0.229
	(0.134)	(0.162)
Per Capita Farmland (V11)	0.0495	0.0209
	(0.0864)	(0.165)
Per Capita Income (V12)	−1.22 × 10−6	−2.23 × 10−6
	(1.91 × 10−6)	(2.19 × 10−6)
Party Member (V13)	−0.143 *	−0.243 **
	(0.0836)	(0.0996)
Local Official (V14)	−0.263 *	−0.129
	(0.139)	(0.171)
Veteran (V15)	0.165	0.168
	(0.103)	(0.126)
Working Elsewhere (V16)	0.0258	0.104
	(0.0786)	(0.0979)
Weifang (C5)	−0.413 **	−0.0282
	(0.164)	(0.199)
Jinan (C6)	−0.503 ***	−0.306 *
	(0.170)	(0.182)
Zhongshan (C8)	0.405 **	0.750 **
	(0.189)	(0.379)
Cutoff point 1 (α1)	−0.0691	−0.467
	(0.299)	(0.351)
Cutoff point 2 (α2)	0.134	−0.231
	(0.300)	(0.352)
Log pseudo likelihood	−1060.1579	−734.54207
Wald chi2	76.36(25)	48.32(26)
Prob > chi2	0.0000	0.0050
Pseudo R2	0.0371	0.0353
Observations	1209	823

Note: *, ** and *** stands for significant respectively at 10%, 5% and 1% level of confidence.

, with the standard errors parenthesized immediately below the corresponding coefficient estimates. The outcomes of the first regression suggest that those farmers who have experienced land requisitions (V1 = 1) tend to prefer stable land ownership. These results indicate that the loss of land ownership and of chance to participate in the land market caused by land requisition leads to a preference for stable land ownership. Experience of land reallocation (V4 = 1) had no significant effect. This may be because land reallocation is conducted within the village according to demographic change; while this leads to changes in land ownership, the changes are more expected than land requisition, thus leading to less worry about the stability of land ownership. Related to this, farmers who have not witnessed any increase in household size (V9 = 0) tended to prefer stable land ownership. The reason may be that, because farmland in China is equally distributed amongst members of the village collective and are adjusted according to change of population, those with decreased household size are more worried about the adjustment of land ownership. For demographic characteristics, the age (V5) and education (V7) of farmers has no significant effect, yet female farmers (V6 = 0) significantly preferred stable land ownership. That may be explained by the fact that women are more vulnerable in terms of land rights, as they are more likely to move due to marriage and other reasons, which makes it difficult for their land rights to be effectively guaranteed [54]. Family characteristics, household size (V8), off-farm income (V10), per capita farmland (V11), and per capita income (V12) have no significant influence, which may indicate that family characteristics are not strongly related to the stability of land ownership. In terms of social relations, there was no significant effect on those respondents whose families included veterans (V15 = 1) and members working elsewhere (V16 = 1); however, those whose families involve party members (V13 = 1) and local officials (V14 = 1) tended to disagree with stable land ownership, albeit at a lower level of statistical significance. In terms of geographical features, farmers from north China’s cities of Weifang (C5 = 1) and Jinan (C6 = 1) appeared to disagree significantly with stable land ownership; this is in contrast with those from the city of Zhongshan (C8 = 1) in the south. These results may be explained by the fact that marketization in the south is of a higher degree than that in the north, and so farmers in the south have a stronger market conception that leads to a preference for stable land ownership

The second regression paid attention to attitudes toward compensation for land requisition, instead of experience of land requisition. It was found that farmers who were unsatisfied with compensation for land requisition (V2 = 1) tended to support stable land ownership. That may because those who were unsatisfied with compensation were more unwilling to undergo land requisition and instead prefered to hold land or to transfer land on market, which all rely on stable land ownership.

The results of other variables were generally consistent with those of the first regression, although the second test identified the change in household size (V9) as the most significant explanatory variable. For those farmers whose families had ceased to expand (V9 = 1), stable land ownership tended to be their strongly preferred policy option. In addition, the second regression also indicated that females (V6 = 0) tended to support stable land ownership, while those with party members in the families (V13 = 1) tended to disagree. Farmers from Zhongshan (C8 = 1) remained significantly more supportive of stable land ownership in contrast to those from Jinan (C6 = 1), although the latter showed less significance compared with the same figure in the first regression.

6. Conclusions and Discussion

6.1. Conclusions

In this research, we discussed the impact of several factors on whether individual farmers support stable land ownership or not. Our findings entailed a series of scientific and empirical implications, as follows:

(1)

The farmers’ dissatisfaction surrounding land requisition was a main driving factor in their support for stable land ownership. However, this proved not to be the primary reason. Rather, the explanatory variable regarding the change in household size (V9) turned out to have a larger magnitude as well as a higher level of significance than farmers’ dissatisfaction with land requisition. Pervasive rural-to-urban migration, combined with the policy of family planning, has meant that many rural families in recent decades have seen a substantial decline in size compared with the multigenerational large farmer households that used to be very typical in the Chinese countryside. Under the HRS, collectively owned farmland is due to be reallocated based on household size every 15 to 30 years, which would clearly defy the interests of those shrinking families, who thus understandably would prefer stable land ownership to the HRS.

(2)

Gender played an important role, according to our findings, in explaining female farmers’ preference for stable land ownership. In China, a woman after a marriage is supposed to leave her own parents and move into her husband’s household. This transition has farmland implications. For the female farmer’s husband, the increase in household size usually entitles him (rather than his wife), on behalf of the household, to more farmland in the next round of reallocation under HRS. However for the female farmer’s pre-marriage family, the decrease in household size may lead to a cut in their farmland quota. The issue is ultimately attributable to the traditional patriarchic economy in the Chinese countryside, where women tend to be more vulnerable in terms of their economic interests.

(3)

The interaction between the state and the market is conceivable in farmers’ day-to-day perceptions about land rights. Farmers are increasingly conscious of their land-related financial interests and seek fairer, market-based compensations to justify their farmland requisitions. The more dissatisfied they are with the compensations, the more strongly they prefer stable land ownership.

Finally, geography has a relatively minor impact on farmers’ preferences for stable land ownership. Of the 12 cities, only Jinan (C6) and Zhongshan (C8) displayed modestly significant coefficient estimates. This demonstrates that farmers in different urban suburbs share similar preferences for stable land ownership. Combined with the descriptive statistics above, the proportion of “agree” and “disagree” regarding stable land ownership was mostly consistent, except for Zhongshan (C8) and Jiangyin (C3). The regression excluded geographic effects by controlling spatial fixed effects. Some research on farmers’ perceptions of property rights security and land tenure stability [55,56], which takes different areas as samples, have reached similar conclusions. It can be proven that regional background has little effect on farmers’ perception of land.

6.2. Further Discussion

From an intellectual perspective, this paper demonstrates a rigorous social survey and statistical analysis approach to systematically study farmers, as opposed to studying the state and the market, which are the two usual subject matters in this kind of research. If stable ownership of farmland can be considered a feature of market-based land use policies, our empirical research simply revealed some important grassroots-based local Chinese characteristics that have been driving the country’s continuing marketization. As local governments are the actual managers of urban land, the existing literature on land marketization has mainly focused on the influence of local governments’ behaviors on urban land marketization [57,58,59]. However, the ultimate goal of land marketization is to build a unified urban and rural land market. As such, present studies on the release of the market value of rural land and the promotion of the rural construction land market is insufficient. Our analysis focuses on farmers’ perceptions of stable ownership and provides some evidence of their support for further marketization. Some studies have proven that clear ownership of rural collective land in some rural areas in China contributes to promoting the price of state-owned industrial land and improving the marketization level of urban and rural construction land [60]. Farmers’ preference for stable land ownership will promote the identification of property rights, which benefits the formation of a rural land market.

In conclusion, we argue that the Chinese version of marketization, as reflected in the farmland domain, involves some unique grassroots-based local Chinese characteristics. As there is a large rural area in China, the study on Chinese land marketization cannot ignore farmers’ perceptions. Those farmers who support stable land ownership tend to be from smaller households, female, and dissatisfied with land requisition. Their preference for stable land ownership is aligned with, though not always determined by, the marketization of the overall economy, of which the land market is an important element. We shall continue to orientate our future research to the farmer grassroots, hopefully supplemented our further study with more in-depth qualitative fieldwork and action research that would analyze the Chinese farmers’ everyday actions in contemporary China’s urban periphery.

Some researchers have mentioned that farmers’ residential ownership and farmland ownership will affect their knowledge about, attitude towards, and practice of rural land changes. In contrast, our paper focused on the impact of land changes on land ownership preference. Farmers’ knowledge and attitude will also affect their preferences. Thus, the interaction of these three factors is subject to further research. There is also an area of the literature that pays attention to the fragmentation of land ownership and discusses its evolution, content, and disadvantages to emphasize the importance of stable land ownership in sustainable land use [61]. This is worthy of further discussion, combined with our study.

However, there are some limitations to our study. For example, we used data collected between 2008 and 2009 for the empirical test, which was not up to date and may have reduced the effectiveness of our findings in explaining the reality. Furthermore, while we concentrated on China’s marketization with the example of land marketization, more examples of cases are needed to gain a more comprehensive and profound knowledge of the role of grassroots in marketization.