How G overnment’s Policy Implementation Methods Influence Urban V illagers’ Acceptance of Urban Revitalization Programs: Evidence from China

: This study aims to provide new knowledge for the governments to enact more effective policies to proceed housing redevelopment programs. We conducted a survey on local urban villagers in Hangzhou city of China. Overall, our results provide valuable theoretical and practical implications for sustainable urban development. Firstly, we found that more reasonable compensation and more respecting justice and democracy during redevelopment implementation increases people’s acceptance of government’s housing redev elopment program. Secondly, we demonstrated that experiences from social learning, such as government – homeowner conflicts and quality of living of other homeowners who have experienced similar programs, and people’s own historical housing redevelopment experience, significantly influence their acceptance of the program. Thirdly, if the governments ensure more justice and democracy, it largely enhances effectiveness of compensation in promoting people’s acceptance of the housing redevelopment programs.


Introduction
Conflicts between government and homeowners in urban revitalization are common in developing countries [1][2][3], and most of the conflicts arise from the process of internalizing externalities which are understudied [4][5][6]. This study aims to provide new knowledge for the governments to enact more effective policies to proceed redevelopment by internalizing the positive externalities of urban revitalization.
We selected the redevelopment of urban villages in the Hangzhou city of China as field of experiment. Rapid urbanization lead to large scale of urban housing demolition. For instance, 68 urban villages were demolished and reconstructed by the end of 2015 in Hangzhou, the capital of Zhejiang Province. The local government further plans to complete the redevelopment of 178 urban villages in the next five years, involving 58,790 households. Long-term and large-scale redevelopment has been common in many countries. For example, the first period of mass demolition in England started from the slum clearance act of 1930 and again as part of the post-war housing reconstruction between the mid-1950s and the 1960s [7,8]. After the Great Depression and World War II, the United States started a large-scale slum clearance program. More than 400,000 housing units were cleared or planned to be cleared, and over 300,000 households were relocated [9][10][11]. Over the last decades, the United States and many European countries have witnessed substantial housing and neighborhood redevelopment programs, such as HOPE VI (USA), urban renewal in France and 'Urban Restructuring' in the Netherlands [12].
An urban village in China is typically a community located in the very central area of a city, but it is dirty, messy, and filled with poor residents [13,14]. Different from the western economies where the slums are the result of urban declining or outdated buildings [15] during the period urbanization is almost down, the urban villages in China are an outcome of China's unique land ownership system during the rapid urbanization process [16]. However, the urban villages in China share similar causes to that of the western economies, which means it is cheaper for the local governments to acquire lands from the sub-urban areas and to build up sub-city centers compared to redeveloping the central outdated urban areas.
We conducted a survey on local urban village households whose homes have been, are being, or are to be redeveloped in Hangzhou, China. We found that, in addition to urban villagers' demographic characteristics, the governments' policy implementation methods significantly influence urban villagers' attitudes towards potential housing redevelopment programs. Firstly, if the compensation is more reasonable, the redevelopment implementation is more standardized, the government respect people's demographic rules more, people will have higher willingness to accept government's redevelopment programs. Secondly, individuals observe and learn from the social issues like government-homeowner conflicts. Thus, people's perceptions on governmenthomeowner conflicts and understandings on other people's quality of living after housing redevelopment significantly influence their acceptance of government's potential housing redevelopment programs. We also find that if a household has positive experience of housing redevelopment in the past, their acceptance of future governments' redevelopment program will be higher. Thirdly, more justice and democracy during land acquisition and redevelopment procedure enhance the effectiveness of compensation in promoting people's acceptance of the redevelopment.
This paper contributes to the urban revitalization literature [4][5][6] by providing more understandings on the process of government's efforts to internalize the externalities. We argue that (monetary) compensation alone is insufficient to persuade homeowners to accept redevelopment programs, but the justice and democracy during redevelopment implementation ensure the effectiveness of compensation in raising people's acceptance of redevelopment. This study differs from existing studies on land acquisition and urban redevelopment in China [1,2,17,18] by revealing the mechanism that people learn from social issues as well as their own past experience which influences their acceptance of governments' redevelopment programs. Moreover, this paper is of policy significance, and it provides reference for governments' implementation of urban revitalization programs to facilitate sustainable urban redevelopment.
The rest of the paper is structured as follows. Section 2 provides a brief literature review. Section 3 introduces the field of experiment, data, and research design. Section 4 elaborates our results and discussion. Section 5 concludes and discusses policy implications.

Urban Revitalization
The revitalization of outdated urban areas can enhance urban living environment and land use efficiency, to raise the overall social welfare. Literature has an extensive discussion on externalities of urban revitalization [19][20][21][22]. The factors influencing the magnitude of externalities include distance from the urban revitalization sites and some government housing policies like rent control [23]. The externality of urban renewal is often measured by economic consequences, like changes in housing prices [15,24].
The governments' effort to initiate and pay for community revitalization is a typical process of internalizing the externalities of urban revitalization [4][5][6][25][26][27]. Internalization of externalities is vital to the sustainability of urban revitalization programs, due to the low motivation of homeowners [5]. One the one hand, owners of the outdated or dilapidated buildings typically are unwilling to participate in state-led upgrading programs. On the other hand, they are often unsatisfied with the compensation from community upgrading programs, on the other hand, they are not able to benefit from (or internalize) the positive externalities of housing revitalization. While the externalities of urban revitalization have been widely discussed, the internalization process of externalities is insufficiently studied [24]. To fill in this gap, our study aims to investigate how the factors, which are related to government's behaviors, influence the internalization of externalities in urban revitalization. We will take the redevelopment of urban villages in Hangzhou China as a field of experiment.

Conflicts between Governments and Homeowners in Urban Redevelopment
Urbanization is a necessary process for the development of an economy [28]. However, as the urbanization process continues to accelerate, the demand for land by capital, technology, and other production factors is increasing, making limited urban land resources scarcer [13,29]. Therefore, on the one hand, cities will inevitably acquire land from rural areas. On the other hand, the city must redevelop its outdated areas to make better use of its existing lands. Conflicts between the governments and the homeowners are very common during the redevelopment process in developing countries. In China, between 2011 and 2015, more than 40% of administrative dispute cases are associated with urban redevelopment [30]. For example, substantial conflict between households and local government resulted in the abortion of Hung Hom Estate redevelopment project in Hong Kong [31]. In the Hongqiao Comprehensive Transport Hub of Shanghai, conflict between homeowners and local government caused violent resistance from the property owners, which resulted in project delay [32].
Due to the rising conflicts between the government and the homeowners, urban redevelopment has been increasingly challenging [2]. Many conflicts between the government and homeowners turn out to be violent resistance of the homeowners, and these are more common during the housing demolition process. Given that the Chinese treasure their housing as an essential private property, they are less willing to give up the property and resettle. With rapid urbanization, recent years have seen increasing and severe resistance to housing demolition from local residents, who are referred to as "nail households" [33]. They refuse to access the government's plan and present strong resistance until they receive the amount of compensation that they consider 'fair' [34]. Once the government refuses to satisfy their requirements, "nail households" will take more extreme actions to confront with government's demolition, and some even risk their lives [35]. To some extent, governmenthomeowner conflict has become an important factor that has deferred China's sustainable urban development, social stability, and efficiency [2].
Since the reform and opening-up in China, the rapid urbanization process has generated a large number of urban redevelopment projects for urban expansion and urban infrastructure improvement [32]. Urban villages are the most commonly targeted areas that are redeveloped [36,37]. The urban villages are dirty, messy, poor, and even regarded as 'China's slum' [14,38]. Given increasing government-homeowner conflicts during large-scale urban village redevelopment in China, it has been a social and academic focus to study what are the causes of and how to address the conflicts [1,17,18]. Many studies argue that compensation is the most important cause for governmenthomeowner conflicts. Intuitively, homeowners are not satisfied with the redevelopment programs either because they are not sufficiently compensated or because they are not aware that they have been fully compensated [1][2][3]. Some studies also argue that the government-homeowner conflicts are also caused by "government's unrestricted power" and imperfect legislation [1,2,39]. For example, Ocheje (2007) criticizes the "public interest justification for forced evictions of African governments" was actually a cover for antiquated and inappropriate planning laws of colonial origin, corruption, and failure of development and land reforms. Similarly, Pils (2005) [3] pointed out that many issues of the demolition are caused by lack of protection for private property in legislation. Therefore, the central government of China and even local governments have carried out a series of adjustments to the land acquisition and demolition policies, strictly regulating the management of land acquisition and demolition [40]. In 2011, the Central Government officially promulgated regulations about the collection and compensation for houses on state-owned lands. Since then, the compensation has been significantly raised, and the grass-roots regime has been disciplined and supervised. However, despite this, conflicts between homeowners and the government continue to occur [2]. The above anomaly gives rise to our research question: what are the factors that affect homeowners' acceptance of urban village redevelopment in China?

Background of Urban Village Redevelopment in Hangzhou
We conducted a survey in 2017 on local urban villagers in Hangzhou, China. Since the reform and opening-up, China's urbanization process has further accelerated. Between 1978 and 2004, the urban built-up area increased from 8,842 km2 to 23,943 km2 [41]. Meanwhile, the proportion of urban citizens in total population rose to 45.7% in 2008, which was more than two times of that in 1978 [42]. To support rapid urbanisation and accommodate rapidly increasing population, cities have expanded rapidly and kept converting surrounding rural lands to urban lands. In this process, the government are more willing to acquire farmland in the sub-urban areas while less willing to redevelop the outdated villages in the central areas to avoid high compensation costs [13]. Consequently, these outdated villages are gradually surrounded by new urban areas and are transformed into the urban villages. Compared with average urban housing, the homes in the urban villages have cheaper rents due to their poor housing conditions and inferior environments. However, the urban villages usually have good access to public transportation. Therefore, the urban villages become the most favourable places of residence for migrant workers and other low-income groups [43].
Homeowners in urban villages have long-term land use rights and can construct housing on their lands free from government urban planning administration, but they are not allowed to sell their properties to residents outside the urban villages or even non-household members [44][45][46]. As a result, the urban villagers tend to substantially expand their housing on the purpose of providing more renting space. In addition, these local villagers have low motivation to well maintain their housing because most of the rooms are used for rental purposes [45]. Consequently, urban villages then have issues like overcrowding and squalor, as well as being full of social problems [13].
However, China's urban villages are different from slums in other countries. As suggested by UN-Habitat, slums refer to those urban residential areas with "inadequate access to safe water, inadequate access to sanitation and other infrastructures; poor structural quality of housing, overcrowding and insecure living conditions". Although urban villages are like slums in terms of poor physical conditions and management, those villages in Chinese cities cannot be classified as slums [38]. There are two aspects of major difference between the urban villages and the slums. Firstly, the residents are different. In the urban villages, the homeowners act as the landlords who live on renting their private properties, and renters are the low-income migrant workers. Urban villages tend to be homes for the new migrants where part of whom will finally move out and upgrade to normal citizens. However, the slums are home to the poor and crime. Secondly, the causes of formation between the two are different. Urban villages are those areas which the governments choose not to develop due to some reasons, but the slums are the declining urban areas without proper maintenance and management [16].
Hangzhou, the capital of Zhejiang Province, is one of the typical cities suffering from urban village problems after rapid urbanization. The redevelopment of urban villages in Hangzhou consists of three modes, demolition and resettlement, regeneration, and the combination of the two. According to "Suggestions on the Redevelopment of Urban Villages in Hangzhou (2016-2020)" which was announced by the Hangzhou Municipal Committee and Municipal General Office, the redevelopment of urban villages in Hangzhou consists of three modes: 1) Demolition and Resettlement; 2) Regeneration; 3) Combination of the two modes. Firstly, in the mode of Demolition and Resettlement, buildings in urban villages would be demolished according to overall urban plans, and local villagers would be compensated and resettled. According to the plan, the majority of urban villages (139 out of 178) will be redeveloped with this mode, which indicates that 79,749 households will be resettled. Secondly, the regeneration mode will be adopted for those urban villages in which infrastructures are relatively completed, and buildings are in good conditions. The regeneration mode focuses on removing illegal buildings and refurbishing buildings' outer facades, to improve the neighborhood environment. According to the plan, 22 urban villages will be redeveloped urban the regeneration mode, influencing 10,613 households. Thirdly, a combination of the above two modes will be adopted in 17 urban villages, involving 11,355 households. The urban village redevelopment studied in this article refers to the mode of demolition and resettlement.

The Survey and Data
Given that Hangzhou is a representative city that has a large number of urban villages having been or to be redeveloped. There were 246 urban villages in total in the main urban area of the city. Redevelopment of 68 urban villages had been completed by the end of 2015. Redevelopment of the rest 178 (as shown in Figure 1) are planned to be completed during 2016-2020. The redevelopment in Hangzhou is in a very large scale. For example, the government planned to start the redevelopment of 25,959 households in 2017 alone. Note: In the figure, each dot describes one urban village. We only consider the main urban area in Hangzhou, which are the areas within the red boundary. We randomly selected and surveyed 20 of the urban villages which are marked by red triangles.

After interviewing relevant government departments like Hangzhou Ministry of Land and
Resources and Hangzhou Construction Committee, we randomly selected 20 urban villages and we conducted the survey between December 2016 to August 2017. In order to increase response rate and reduce selection bias, seven research assistants were hired to approach the potential respondents face to face. Then, 330 valid questionnaires were collected (each questionnaire is answered by one representative of one household). The sample size meets the requirement (the standard deviation is 0.45762. At 95% confidence level, the minimum sample size is 322, with the estimated error no more than 5%).
The sample distribution is presented in Table 1. Among our sample of 330 households, a total of 160 households (one representative for each household) had experienced redevelopment, accounting for 48.5% of the total sample. 121 households' homes are being redeveloped (36.7%) and 49 households' homes are pending to be redeveloped (14.9% of the sample). In Panel B of Table 1, redevelopment is welcomed by 68.97% of interviewed households. Table 2 presents the distribution of demographic information of the respondents.  ) and their willingness to accept another government redevelopment program in the future, which is indicated as "Willingness to be redeveloped" (Panel B), which is the key dependent variable in this study. In Panel A, "Redevelopment finished" means the household has already experienced redevelopment and the redevelopment has been finished; "To be redeveloped" means that a household has reached agreement with government and its home is planned to be redeveloped. "Being redeveloped" means the demolition of a household's home has already started.  Total  100  Total  100  Total  100  Total  100  Total  100  Total  100 Note: This table provides the distribution of the surveyed household representatives in terms of each demographic factor. In Column 3, the category "Single" includes these unmarried and the divorced.

Model Specification
Our empirical analysis aims to explore the factors that affect attitudes of homeowners in urban villages towards housing redevelopment. We adopted the binary logistic regression model, which is specified in Equation 1.
is the dependent variable. Y i =1 if the respondent is willing to accept the future housing redevelopment program conducted by the government, and Y i =0 if he/she is unwilling to accept the redevelopment program.
represents a vector of explanatory variables which are discussed in Section 3.4.
is the vector of coefficients. is the residual.

Variable Selection and Description
The dependent variable, "Willingness to be redeveloped", is measured by the question "Given that the government will provide reasonable compensation level, are you willing to accept government redevelopment plans in the future?". The dependent variable measures people's perception on government's redevelopment plans. A household's willingness for redevelopment depends on the rational judgment about benefits from redevelopment. We divide the factors influencing households' judgement into three categories.
The first category is individual characteristics and household resource. This category includes individual demographic and household characteristics, such as gender, education, marriage, and household income. Besides, household members working as government staff were chosen as an important factor of household resource in the social structure, which represents access to more social capital. Social capital helps regulate the rationality of household behaviors. Therefore, the existence of household members working as government staff can encourage the willingness to accept land acquisition and redevelopment. Besides, as a social stabilizer, social security system can effectively alleviate structural pressure and thus avoid the occurrence of destructive behaviors of residents. Since the urban and rural medical insurance has basically covered the whole area of Hangzhou, this study will choose participating in urban pension as the main indicator for whether a household is protected by social security.
The second category is perception of society. Previous events continue to affect people's understandings about society and people's values, thereby ultimately influencing their attitudes towards urban village redevelopment. Past events, such as conflicts in land acquisition and redevelopment and corruption of government officials, may all be sources for people to learn and understand society. During the period of rapid social transformation, changes in people's land awareness show that this type of impact is significant and important. Chinese farmers' attitudes towards land acquisition and redevelopment have changed significantly in recent years from supporting and trusting the government and fully accepting the compensation and resettlement policies to gradually safeguarding their rights and trying their best to resist [45,47].
Therefore, we used the following questions to dig out people's perception towards to society. We use the question "Do you think that the conflicts between the government and homeowners are evident in land acquisition in China?" to measure people's understandings on whether "(Government-homeowner) conflict is common". We used the question "Do you think that many forced redevelopment events have occurred?" to measure people's understandings on whether "(Unwillingly) forced eviction/demolition is common". We used the question "What do you think of the situation of people after land acquisition and redevelopment" to measure people's perception of quality of living after redevelopment.
The third category is the perception on justice in the process of land requisition and redevelopment. We used the question "During the process of land acquisition and redevelopment, do you think the compensation is reasonable?" to reflect people's perception on compensation level.
We used the question "Do you think that the process of land acquisition and redevelopment is compliant with rules?" to measure people's perception on whether the redevelopment process is compliant with rules. We used the question "Do you think that the process of land acquisition and redevelopment respects people's democratic rights?" to reflect people's understandings on government's respect for citizen's democratic rights. Table 3 summarizes the definitions of the variables discussed above. Table 4 provides the distribution of each variable.

Urban Villagers' Willingness for Redevelopment
According to our survey, as shown in Table 5, 68.97% of all the interviewees indicate that they are willing to accept future government redevelopment programs by the government, and 31.03% refuse. Among the respondents who have reached agreement with government for the current redevelopment plan, 76.5% of them show their willingness for future redevelopment. The number is higher than that of the group of households who have already experienced housing redevelopment (63.1%) and that of the group of households whose homes are being redeveloped (69.1%). The main reasons why people will accept future redevelopment include "acquiring more than one set of resettlement housing units to improve living environment and rental income", with 72% of the interviewees who are willing to accept future redevelopment choosing this; "Obtaining compensation to improve quality of living", with 66.29% of the interviewees who are willing to accept future redevelopment choosing this; "Improving infrastructure and public service conditions such as transportation, schools, communications, and medical services", with 34% of the interviewees who are willing to accept future redevelopment choosing this. Note: This table provides the attitudes to a future redevelopment program of three groups of residents, the "Redevelopment finished" group, the "To be redeveloped" group and the "Being redeveloped" group. "Redevelopment finished" means the household has already experienced redevelopment and the redevelopment has been finished; "To be redeveloped" means that a household has reached agreement with government and its home is planned to be redeveloped. "Being redeveloped" means the demolition of a household's home has already started. From Table 5, we can see that people's perceptions on future housing redevelopment are generally positive. People look forward to government redevelopment as they believe redevelopment is beneficial. In fact, with increasing value of compensation for lands and houses due to urbanization, most homeowners not only do not reject the redevelopment, but also expect to change the existing economic conditions through land acquisition and house redevelopment, given reasonable compensation (as shown in Table 6). Shi et al. (2011) finds that land acquisition and redevelopment may not reduce household income and may even increase household income in developed regions [48]. However, we observed some anomalies and concerns: Why is the perception of people towards future redevelopment positive, but it turns to be less positive after people's housing redevelopment starts? Are there other factors that fundamentally affect people's perception for redevelopment, in addition to economic factors like compensation? An in-depth analysis is needed to answer these questions to provide a better way to resolving current redevelopment difficulties.

Empirical Results: What Influences People's Perception on Urban Village Redevelopment
A logit model (Equation 1) was adopted to estimate Models 1-3 and the results are presented in Table 7. The models are of good explanatory power. From Model 1 to Model 3, more explanatory variables are included, prediction accuracy of the model increases, and the coefficients are stable. Therefore, our empirical analysis only needs to discuss the results of Model 3 (Nagelkerke R 2 = 0.510, with a forecast accuracy of 78.6%).
(1) Individual demographic and household characteristics. Among the seven variables of individual and household characteristics, all variables, except for gender and participating in urban pension, are significant at 5% level or 1% level. Thus, factors such as age, educational background, household income, and member of household as government staff, have significant influence on people's perception on future urban housing redevelopment. The coefficient of age is −0.17, thereby indicating that the younger the respondents are, the more likely they are willing to be redeveloped; that is, the less likely they are to reject redevelopment programs. It is traditional in Chinese culture that people have strong community attachments, and they prefer to stay in the houses in which their ancestors lived. This geographical attachment tends to increase with age. Consequently, the older residents' intention for redevelopment is less positive. This conclusion is consistent with our expectation and with previous literature [49]. Among the three categories of the variable of marriage, since the total compensation is directly linked to the number of household members, married households are more willing to be redeveloped than singles. In addition, the coefficient of household with children (1.51) is greater than that without children (0.61). Due to government's redevelopment compensation policy, people in areas of great redevelopment potential are eager to get married and give birth to children to obtain more redevelopment compensation by increasing population. This behavior has also increased the instability of marriage to some extent. In recent years, the divorce rate in Hangzhou has continued to rise, and 27,470 pairs divorced in 2017, a significant increase of 17.6% over the previous year.
Moreover, a positive coefficient of education (0.40) indicates that a high probability of accepting redevelopment is associated with higher level of education. With higher education levels, people have a higher capability to understand the details of the redevelopment policies, and they can also better understand the social values (positive externalities) of urban housing redevelopment and its significance toward improving production and living environment. Household income (−0.110) and household members as government staff (0.742) have also a significant effect on the redevelopment perception. The results show that people with higher income, who are a little sensitive to the economic benefits of the redevelopment, are reluctant to be relocated. Households with civil servants are more willing to be relocated compared with those without.
(2) Policy perception. In line with our expectations, in addition to perception on compensation level (0.533), other variables, such as redevelopment's compliance with rules (0.16) and respect for democratic rights (0.528), also have significant effects on attitudes towards future redevelopment. If the compensation is more reasonable, the redevelopment implementation is more compliant with rules, the government respect people's demographic rules more, and people will have higher willingness to accept government's redevelopment programs. Several studies have proposed policy recommendations to protect the rights and interests of the land-deprived households from various perspectives, such as compensation level and democracy. However, it is also possible that people's feeling of low compensation is caused by their unreasonably high expectation. Wang and Lu (2013) surveyed 493 respondents from 23 communities that are under redevelopment or to be redeveloped in Huangshi city in China and found that approximately 40% of the households require compensation at over 120% of the original housing value [50]. Zhang and Zhang (2013) also found the same phenomenon. Some of the land-deprived households consider land acquisition as a shortcut to being rich and ask for unreasonable compensation [51]. Given such high expected references, households' understanding of "reasonable" compensation might be higher than the reasonable compensation in practice. Therefore, the government should establish a relatively sound and justified compensation mechanism. The model results demonstrate that, in addition to the absolute value of redevelopment compensation, government should also focus on solving the redevelopment implementation issues related to people's perception on "Redevelopment's compliance with rules" and "Respect for democratic rights".
(3) Perception on society. Individuals learn from the society and adjust their behavior following the society. Our results in Table 7 show that people's perception on "(Government-homeowner) conflict is common" and "Perception of quality of living after redevelopment" significantly influence their perception on acceptance of future redevelopment. However, the perception on "(Unwillingly) forced eviction/demolition is common" is insignificant. This is because forced redevelopments have been forbidden in China. As many as 92.3% affirmed that forced redevelopments in Hangzhou are rare. Therefore, to promote harmonious urban village redevelopment, positive public opinions towards government must be built up by the government.

Impacts of Redevelopment Experience on Willingness for Redevelopment
We further divided the survey sample into two groups, the group of respondents that have experienced redevelopment (ER group), and the group of respondents that have not experienced redevelopment (the NER group), to conduct a heterogeneity test. The results are provided in Table 8. For the ER group, all variables under "Individual demographic and household characteristics" category are not significant, except Education. All variables under "Perception on Society" are not significant, while variables under "Policy perception" are significant. Specifically, the coefficient of "Perception on compensation level" is 0.768 and significant at 1% level. It is substantially higher than that of the NER group (which is 0.385). However, the coefficients of "Redevelopment's compliance with rules" and "Respect for democratic rights" (which are 0.141 and 0.475, respectively) are lower than those of the NER group (0.308 and 0.658).
To further explore the reasons for the difference between the coefficients of the two groups, a time dummy as well as interaction term with the "Perception on compensation level" was added to the empirical model using ER group. Time=1 if the year of redevelopment experienced by the respondent is in or after 2011, and Time=0 if the year of redevelopment experienced is before 2011. Since 2011, "Regulations about the Collection and Compensation of Houses on State-owned Lands" takes effect and it legalizes that the homeowners' rights should be better protected, and homeowners should be sufficiently compensated in redevelopment. In addition, for both ER and NER groups, we also included the interaction terms "Perception on compensation level×Redevelopment's compliance with rules" and "Perception on compensation level×Respect for democratic rights", the results are in Table 9. For the ER group, the coefficient of "Time" is 0.39 and significant at 1% level, which suggests that with the implementation of regulations, redevelopment leaves a better impression on the redeveloped households and make them have more positive perceptions on urban redevelopment. The coefficient of "Perception on compensation level×Time" is significantly negative, it implies that the regulation has reduced the impact of compensation level on the people's perception on future redevelopment, which means that people are less worried about insufficient compensation. The coefficients for interaction terms "Perception on compensation level×Redevelopment's compliance with rules" and "Perception on compensation level×Respect for democratic rights" for both ER and NER groups are positive and significant, which means that if the redevelopment is implemented compliantly with the rules and the democratic rights of homeowners are respected by the government, sufficient redevelopment compensation level can better raises the people's willingness to accept future redevelopment.
This also explains why in developed areas where the level of compensation has been relatively high and redevelopment is generally welcome, households' willingness to be redeveloped in future will be less positive if they have experienced redevelopment or are experiencing redevelopment. The main reason is that the redevelopment procedures and democracy have not met people's expectation. Therefore, besides focusing on the absolute value of compensation and establishing a relatively reasonable and fair compensation mechanism, it is also important to enhance the justice and democracy during land acquisition and redevelopment. Social psychologist Taylor argues that if people believe that a result is produced through a fair and standardized process, they are more willing to accept the outcome whatever it is. In other words, the fairer the procedure is, the easier people will accept the outcome of the procedure. Liu et al. (2012) concluded that procedural justice is more important than monetary compensation in determining farmers' satisfaction with land acquisition in China [40]. Note: This table compares the empirical results for the sub sample of residents who have experienced housing redevelopment program to the results for the sub sample of residents who have not experienced housing redevelopment program, to reveal how previous redevelopment experience influence residents' willingness to accept a future redevelopment program. The results focus more on how government's method of redevelopment implementation influences the impacts of compensation level on residents' willingness to accept housing redevelopment program. *** p<0.01,** p<0.05,* p<0.10.

Conclusions and Implications
We conducted a survey on urban village households whose homes have been, are being, or to be redeveloped in Hangzhou, China. We found that, in addition to the individual demographic and household characteristics, government's implementation of redevelopment policies significantly influences urban villagers' acceptance of the redevelopment program. This leads to the following implications.
Firstly, if the compensation is more reasonable, redevelopment implementation is more compliant with rules and the governments respect people's demographic rights more, people will have higher willingness to accept government's redevelopment programs. Furthermore, if more justice and democracy is ensured, it largely enhances effectiveness of compensation in motivating people to accept housing redevelopment programs. This indicates that in addition to monetary compensation, establishment of a democratic, explicit, and standardized procedure should be one important part of policy reform for land acquisition and demolition. This requires the governments to optimize and clarify every aspect of the procedure, including purpose, executor, public consultation, compensation plan, social stability risk assessment, judicial enforcement, and disclosure. More importantly, public participation in the redevelopment decision making is necessary in an explicit system, to ensure a fair and transparent redevelopment process and to protect people's right of being informed and participation. People are more willing to accept and comply with the decision results if the procedure is fair, democratic, and normative.
Secondly, individuals learn from social experiences (what he/she observes, hears, reads, etc.) and adjust behaviors accordingly. People's perceptions on "(government-homeowner) conflict" and "quality of living after redevelopment" significantly influence their willingness to accept future redevelopment programs. People's perceptions and assessments of past land acquisition and demolition converge to public opinion and social atmosphere, which further influences their behavior and perceptions again, forming a feedback loop which amplifies the shock that issues in government's redevelopment programs have on people's perception towards the future redevelopment program. Social media is an important channel for people to gain social experiences. In this regard, the social media should be guided by the governments to provide objective information about social issues like government-homeowner conflicts, which guarantees a healthy social atmosphere and objective public opinions.
This research is very preliminary research on the channel of social learning and experience learning through which people form perceptions on government's housing redevelopment programs. Limited by time and resources, our questionnaire design does not allow us to conduct an in-depth analysis on what else we can do, in addition to guiding the social media, to intervene the channel towards healthy social atmosphere and objective public opinions.