Policy Implications and Impact of Household Registration System on Peasants ’ Willingness to Return Rural Residential Lands : Evidence from Household Survey in Rural China

Despite a growing body of literature on China’s household registration system and rural land transfer, few studies have examined the impact of the household registration system on peasants’ willingness to return rural residential land. This paper aims to fill this gap and uses household survey data to measure the impacts of household registration system on peasants’ willingness to return rural residential land. The results show that the household registration system reduced the farmers’ enthusiasm to exit the rural residential land, that is, household registration system had a significant negative impact on farmers’ willingness to return rural residential land.

Contradictions between the urban-rural population and land use structure have had profound impacts on social and economic development in transitional China.With the gradual increase of urbanization, much of the rural surplus labor force has moved to the cities in the pursuit of new lives (Hualou Long, Jian Zou, and Yansui Liu 2009).Most of these laborers began to work and live in towns and cities but retained their residential land in rural areas, which caused many hollowed villages to form (Liu et al. 2010).Although the rapid development of urbanization and industrialization in China have reduced the total area of rural land, residential land appears to be an increasing trend in most of these rural areas of China (Hui Wang, Ran Tao, and Lanlan Wang 2012).This large disparity has captured public attention and caused much discussion.The government has taken measures to address the problem, but the practical effect of the policy is not ideal.The household registration (hukou) system, which reshaped the dualistic (urban vs. rural) structure of Chinese society, is one of the main reasons for these changes.A large and increasing number of rural people are moving to urban areas but are not able to convert their hukou from rural to urban and are thus subject to institutional discrimination against rural hukou holders.Every rural household is eligible to apply for only one parcel of residential land for free, but once the farmers give up their rural status, they retain no right to free rural residential land.Thus, from the theoretical point of view, the household registration system influences the peasants' intentions to leave their homesteads, but the issue of the peasants' willingness to return rural residential requires further study.This issue also relates to improvements in land use efficiency and farmland protection and can help farmers obtain property income and reduce the costs of entering the township.

Literature Review
With the reduction in difference between urban and rural land use in China in recent years, the transfer of rural residential land has attracted the widespread attention of government and academics (Jing Zhou and Qingyuan Yang 2012;Božo Stojanović 2013).Many scholars are currently focusing significant attention on the legitimacy of rural residential land transfers, their social and legal constraints, and institutional innovation, usually from a macro level (Brian Schwarzwalder and Prosterman Roy 2002;Margo Rosato-Stevens 2008;Jun Wen 2010;Benjamin Barros 2012).Households, in contrast, may be seen as the micro-level behavioral agent of investment, production and consumption in rural areas and as the basic decision-making unit of rural land use.The household not only represents the economic rational man but is also the social rational man (Anyck Dauphina and Bernard Fortin 2001).Thus, in the decision-making process of returning to residential land, the household is not simply seeking maximum economic profit but also considering future sustainable livelihood and attempting to maximize combined economic, social, and survival benefits.Some of the literature has accordingly begun to pay attention to the micro-level issues of returning rural residential land by analyzing the impacts of several factors, including personal preferences, farm holder attitudes, and socio-economic factors such as household composition, age, educational level, and social capital, on peasants' willingness to return rural residential lands (Roel Jongeneel, Polman Nico, and Slangen Louis 2008;Yiran Zhang and Daochi Qiu 2011;Kathryn Williams and Jacki Schirmer 2012;Juan Ramón Murua, Inmaculada Astorkiza, and Begoña Eguía 2013).
Similar to other developing countries, China's massive labor migration is mainly driven by the rural-urban income gap.Despite the vast scale of migration in China, there is evidence that constraints on labor mobility remain.The literature describing the sources of these obstacles to migration has mainly focused on the potential destinations of migrants.The household registration system (known in China as the hukou system) has been described as the major impediment to migration because it prevents rural migrants from accessing all of the benefits associated with legal residence in cities (Katrina Mullan, Pauline Grosjean, and Andreas Kontoleon 2011).Individual hukou defines for each Chinese citizen a set of rights and opportunities and the localities in which they can be exercised (Thomas Vendrys 2011).This institution thus has a strong influence on individual locational choices.Beginning in the late 1990s, the authority to grant urban hukou has been handed over from the central government to local governments.Many cities have also taken measures to eliminate the distinction between rural and urban hukou within the individual city.However, PANOECONOMICUS, 2016, Vol.63, Issue 1, pp. 135-146 local authorities usually grant urban hukou only to the rich, the well-educated, or those who are immediate family members of current urban citizens.Moreover, the equalization policy adopted by many urban governments only applies to those who have held local rural hukou.Therefore, reforms of the household registration system seem to be irrelevant for most of rural migrants from other regions, and even now, hukou remains an obstacle to rural migrants who live in cities and plays an important role in deciding people's social benefits, such as medical treatment, social welfare, housing, and children's education (Huafeng Zhang 2010;Mingjie Sun and Cindy Fan 2011;Zhiming Cheng, Feng Guo, and Graeme Hugo 2013).
Although there is a growing body of literature (such as Alan Brauw and Rozelle Scott 2008; Yu Zhu and Wenzhe Chen 2010; Ying Xu, Boxin Tang, and Edwin H. W. Chan 2011) on China's household registration system and rural land transfer in recent years, most studies in China focus exclusively on the impact of the household registration system on temporary migration and socio-economic wellbeing.However, hardly any studies examine the impact of the household registration system on peasants' willingness to return rural residential land.This paper aims to fill this gap by using household survey data to measure the impacts of the household registration system on peasants' willingness to return rural residential land.The rest of the paper is organized as follows: the next section presents the analytical framework used to test our research hypotheses; the following section addresses econometric methods and data description; the fourth section discusses the empirical results; and the last section draws conclusions.

Theoretical Analysis and Hypothesis
According to the theory of rational choice, decision-making behavior depends on the expected future results produced by current actions (James March and Johan Olsen 1989).The farmer is presumed to be the "rational economic man", thus reflecting a basic hypothesis of household economic decision-making behavior in transition China.Within the present dual structure of urban-rural, farmers with rural hukou can obtain free access to the use rights of rural residential land.In rural China, the land use system guarantees the farmers' rights to residence and survival by granting famers the right of rural residential land use.Thus, if rural contracted land has a security function by providing the multitudes with food to eat, then the rural residential land in fact has a security function by providing everyone in rural areas with a house to live in.Under the incomplete conditions of the rural social security system, rural residential land also provides old-age security and can be considered the invisible social welfare of the rural household registration system.Meanwhile, according to the provisions of the land management law in China, collectively owned land is limited to the exchange of contract rights among villagers within land tenure (Oskam Arie 2007).Rural land use rights, including rural residential land, cannot be traded in the market.To avoid losing their dwelling places and facing the threat of displacement, farmers may transfer rural residential land (Zhonghua Huang and Xuejun Du 2012).Therefore, the security provided by rural residential land is the key factor affecting the farmer's decision to withdraw from these areas.
PANOECONOMICUS, 2016, Vol. 63, Issue 1, pp. 135-146 Farmers' willingness to return residential land has a fundamental influence on regional economic and social development.For rural farmers, the purpose of returning rural residential land is to seek maximum economic and social benefits.These benefits are based on the farmers' subjective evaluation of ability of the withdrawal behavior to fulfill their needs (Zhaolin Wang and Qingyuan Yang 2011).However, the household registration system assigns the agricultural or non-agricultural status to each child at birth, and hukou status was and, to a large extent, still is a very strong determinant of the rights and privileges affecting socio-economic wellbeing.For example, one had to have local residence rights and often a non-agricultural (urban) hukou to qualify for medical care, unemployment, and retirement benefits; to enroll one's children in school; to obtain a non-menial job; or to qualify for public housing.Most forms of insurance were unavailable to those with agricultural (rural) hukou, and both medical care and education were inferior.Most farmers are still considered rural residents under current China's household registration system, though they may have lived in cities for many years.Rural migrants have little safety net beyond small plots of family farmland and the availability of free rural residential land.The past three decades have witnessed a drastic increase in the number of temporary non-local hukou residents (both registered and non-registered) in urban centers.However, these migrants are not entitled to urban benefits unless they are able to convert to a full urban hukou.Rural-urban hukou conversion is possible but can only occur through very limited channels.Rational farmers usually make a cost-benefit evaluation while deciding whether to return residential land, and only when the risk is in an affordable range and the total income of returning to residential land is greater than the total cost will the quit intentions of farmers change to withdrawal behaviors.Farmers will only feel the benefits of withdrawing once they receive support such as housing, medical care, employment, and education, which will require them to enter the urban household registration system.Under the current institutional arrangement of the dual-structure household registration system, it is impossible for rural famers to receive the public services only enjoyed by urban residents.In contrast, farmers with rural household registration can obtain and use rural residential land for free.This type of welfare decreases farmers' willingness to return residential land.
Based on the aforementioned theoretical analysis, we hypothesize that the household registration system has a negative impact on farmers' willingness to return residential land.

Model Specification
The willingness of farmers to return rural residential land is affected by various factors, including individual and family characteristics and the household registration system.However, we cannot quantitatively observe the willingness of this behavior: according to field research on the willingness of returning rural residential land for each household, the only data sets available apply 0 in case of "don't quit", 1 for "want to quit", or 2 for "uncertainty".The variable data used in this paper are discrete values, and the dependent variables are polytomous ordinal response variables.
PANOECONOMICUS, 2016, Vol.63, Issue 1, pp. 135-146 According to studies undertaken by economists modeling individual-level survey responses on household willingness, an ordered probit analysis would be the suitable choice.As opposed to the probit model, the ordered probit model shows us not only the factors influencing the "what" question but also the degree of influence of each factor.This information provides more effective support for related system reform.Ordered probit models have come into wide use as a framework for analyzing such a response (Eric Lesaffre and Herry Kaufmann 1992).
In modeling, we assume the willingness of farmers to return residential land: y i is the unobserved (latent) variable and is expressed in the following equation: where x i are explanatory variables such as age, educational level, household characteristics, and household registration factors; β i are unknown parameters; and u i is a random disturbance.We also assume that u i is normally distributed across observations.As mentioned above, y i is unobserved.We can only observe whether the willingness of a household is "0", "1", or "2", so what was observed is as follows: where  represents the unknown parameters to be estimated with β i .

Variable Selection and Explanation
According to the existing literature and current research, we think that the influential factors of withdrawing from rural residential land include five main categories: individual characteristics, family characteristics, factors of social security, exit compensation, and the household registration system.Farmer characteristics were represented by gender, age, level of education, professional training or certification, and migrant work time.These factors will have different impacts on farmer's willingness to return residential land.Theoretically, as the head of household, the male often bears the burden of caring for parents and children; thus, they pay more attention to the security function of rural residential land.The older the head of household is, the smaller the chance of engaging in nonagricultural employment is because they are less inclined to withdraw from the rural residential land.Farmers with higher education are more likely to engage in nonagricultural employment activity and prefer to return the rural residential land.Farmers with professional training and long-time migrant working experiences are more willing to exit rural residential land.
Household characteristics were represented by the family sustenance coefficient, thus reflecting the proportion of agricultural income and rates of joining the New Cooperative Medical System (NCMS).Generally speaking, higher coefficients of family sustenance correspond to heavier family burdens, so the coefficient of family sustenance will have a negative effect on a farmer's willingness to return residential land.The proportion of agricultural income is expected to have a negative rela-tionship to a farmer's willingness to return residential land because a greater proportion of agricultural income implies that a family's livelihood relies more upon agricultural land resources.Joining the New Cooperative Medical System is expected to have positive effect on farmer's willingness to return residential land because it implies that the household has attained greater social security.
Aside from individual and household characteristics, other factors affect farmers' willingness to return residential land, e.g., the method of compensation for exiting, the comprehensiveness of the homestead policy, and the impact of the homestead on elder housing security.
The influence of location should likewise not be neglected.In our analysis, this variable is represented by distance from homestead to township.Theoretically, rural locations nearer to townships have better economic development and more similar life philosophy and habits.Therefore, this variable is expected to have a negative relationship to farmers' willingness to return residential land.
The main concern of this paper is the impact of the household registration system on peasants' willingness to return rural residential lands.According to the above theoretical framework, the household registration system has a negative impact on farmers' willingness to return residential land.In our paper, we describe both the role of the household registration system through the farmers' understanding and the impact of the household registration system on peasants' willingness to return rural residential lands.For perceptions of the function of the household registration system, the number 3 indicates the functions of employment, housing, social security and education, the number 2 indicates the functions of social security and education, and the number 1 indicates that household registration system has little effects.For the impact of the household registration system on peasants' willingness to return rural residential lands, 4 indicates a response of "very important", 3 signifies "important", 2 corresponds to "moderate", and 1 means "no effect".

Data and Descriptive Statistics
We obtained the data for this study from rural household surveys conducted in Linqing Municipality in Shandong Province from July to August 2011.A stratified sampling method and face-to-face surveys were used to identify and collect information from a representative sample of over 360 rural households.According to the geographical conditions, natural resources, levels of economic development, and urbanization rates of every town, we first selected four townships in Linqing municipality as the survey sites: Daxinzhuang, Kuangzhuang, Bachalu and Panzhuang.Then, we randomly sampled two or three villages from each of the four townships.The number of samples in each township was determined primarily by the proportion of household size.In addition, some of the sampled interviewees either could not respond or could not respond properly: a total of 317 valid questionnaires were collected (360 issued), resulting in a response rate of 88.06%.We have full confidence in the integrity of the primary sources because we conducted the fieldwork ourselves and visited each household in every village.
The questionnaire included sections on basic household information, perceptions of homestead policy, and farmers' understandings of the household registration PANOECONOMICUS, 2016, Vol.63, Issue 1, pp. 135-146 system.A full description of the explanatory variables in conjunction with their corresponding summary statistics is reported in Table 1.

Estimated Results and Analysis
The empirical results obtained from the ordered probit model estimation are summarized in Table 2.According to Table 2, the ordered probit model results in fairly high goodnessof-fit indices.
According to our results, the estimated coefficient for the perception of the function of the household registration system was negative and statistically significant at the 5% level.This result suggests that, keeping other parameters constant, this perception has a significant negative impact on the farmers' willingness to return residential land.A regression coefficient of 0.873 implies that if this variable improves by one percentage point, the willingness of withdrawing from rural residential land will decrease by 87.3%.A possible explanation for this result is that under the current institutional arrangement of the dual-structure household registration system, the existence of urban household registration makes it impossible for farmers to access the public services only enjoyed by urban residents.However, with rural household registration, farmers can obtain and use rural residential land for free.This type of welfare will decrease the farmers' willingness to return residential land.The variable corresponding to the impact of dual-structure household registration system was also found to have a significant negative impact on farmer's willingness to return residential land.It also passes the significance test at the 10% level.The regression coefficient is 0.692, thus implying that if this variable improves by one percentage point, the willingness of withdrawing from rural residential land will decrease by 69.2%.This variable mainly reflects the negative impact of the household registration system on the willingness to return residential land from the perspective of farmers' subjective perceptions.
The regression results of other significant control variables are represented in Table 2.Among the individual characteristics variables, the estimated coefficient of age was positive and statistically significant at the 5% level.This result suggests that, keeping other parameters constant, the variable of age has a significant positive impact on the farmer's willingness to return residential land and implies that older farmers have less opportunity and ability for non-agricultural employment and therefore pay more attention to the social security function of housing (Zhaolin Wang and Qingyuan Yang 2011).The estimated coefficient of professional training or certification was positive and statistically significant at the 10% level.This result suggests PANOECONOMICUS, 2016, Vol.63, Issue 1, pp. 135-146 that, keeping other parameters constant, professional training or certification has a significant positive impact on farmers' willingness to return residential land.However, the variable of migrant work time was found to have a significant and negative impact on farmers' willingness to return residential land; further, it passes the significance test at the 10% level but is not consistent with the direction of the desired effect.A possible explanation is that farmers with longer migrant work time may feel excluded from the urban household registration system because of rising real estate prices and therefore pay more attention to the function and value of their rural homesteads.
With respect to the effect of family characteristics, only the proportion of agricultural income variable was found to have a significant positive impact on the farmers' willingness to return residential land and passes the significance test at the 10% level.More non-agricultural income for farmers corresponds to higher levels of welfare and, therefore, less willingness to return residential land.This outcome is consistent with the findings of Jianghua Guan, Huangchao Xi, and Huyin Gen (2013).
The mode of compensation for exit and the impact of the homestead on oldage housing security both positively affected the farmers' willingness to return residential land and passed the significance test at the 1% and 10% levels, respectively.This result indicates that while withdrawing from the rural residential land, farmers gave more consideration to future housing security.Therefore, improvements in this area can strengthen farmers' willingness to return residential land.
The location variable was found to be statistically significant at the 10% level, which is consistent with the direction of the desired effect and the findings of Nie (2013).Because a closer proximity of the township to the rural residential land corresponds to higher potential economic value of the rural residential land, farmers in these areas are less inclined to withdraw from their rural residential land.

Conclusions and Policy Implications
The household registration system has important implications for farmers' willingness to return residential land.Applying the ordered probit model and using household survey data, this paper investigates the effects of the household registration system on farmers' willingness to return residential land.The empirical results showed that because of the dual urban-rural system and its affiliated welfare functions, the household registration system reduced the enthusiasm of farmers to exit rural residential land and had a negative impact on farmers' willingness to return residential land.
In addition, variables such as age, professional training or certification, compensation modes for exit, effects of homestead on old-age housing security, and distance from homestead to township were all found to have significant and positive impacts on farmers' willingness to return residential land.However, migrant work time, the proportion of agricultural income, perceptions of the functions of the household registration system, and the impact of the dual-structure household registration system all had significant and negative effects on farmers' willingness to return rural residential land.
Based on the above conclusions, corresponding policy implications are as follows.First, in the process of promoting urban-rural integration and encouraging farmers to return rural residential land, the reform of relevant ancillary systems, especially the urban-rural dual structure household registration system, must be performed at the same time.We must restore the basic functions of the household registration system for civil registration and vital statistics as quickly as possible and, thus, remove the connected social functions.Meanwhile, the government should reform the household registration system and implement a household registration policy with different eligibility requirements for people of different conditions, progressively granting urban residency to rural migrant farmers and their families who are both willing and able to stay in cities and towns where they have had jobs or done business for a long time.A residence permit system should be introduced for rural people in urban areas who have not yet gained urban residency.Basic public services should be steadily extended to fully cover the permanent population of cities and towns so the rural people who live in them can contribute to the development of modern urban life and enjoy it together with the urban people.
Second, efforts should be made to accelerate the establishment and completion of the social security system and further weaken the old-age security function of rural residential land.Current rural residential land circulation mechanism such as LUTRG and central villages unavoidably bring about disputes and injure the household welfare situation of peasants.Farmers with low willingness to return their land are worried about the future: livelihood, occupation, children's education, and pensions.To break these barriers to employment for farmers and ensure they can obtain more jobs in urban areas and earn stable economic income, China must establish a unified urban and rural labor market, improve the farmers' employment environment, and strengthen their professional skills.These improvements will resolve many worries connected to social security, health care, and pensions and improve their willingness to return rural residential land.

Table 1
Descriptive Statistics of Explanatory Variables Source: Calculated based on household survey data.

Table 2
Estimated Results for Ordered Probit Source: Calculated based on Eviews software.