The population flow between cities not only reflects the choice of the urban population who voted with their feet based on different development levels of urban economy, society, and public services, but also exerts a direct bearing on the success of rational and orderly movement of population, coordinated development of large, medium and small cities and the new-type urbanization strategies. Therefore, it is of practical significance to explore the spatial flowing laws and driving mechanism of urban-to-urban population flows in China for realizing rational and orderly distribution of population and the new-type urbanization strategy. In the context of the increasing amounts of urban-to-urban floating population in China, this paper discusses the results, processes, and causes of the spatial mobility of floating population in urban areas based on data of population census and dynamic monitoring of floating population. In addition, it studies the spatial distribution, flow trajectory and forming mechanism of urban-to-urban floating population. This paper uses Arcgis to visualize the spatial distribution and main flow trajectories of inter-cities floating population, then uses regression equations to analyze the main factors driving population movement between cities. The results of this paper indicate that the urban-to-urban floating population tends to flow to developed regions and metropolises, during the process of which the proximity of destination is considered; gaps among regions in terms of economic development, employment opportunities, and high levels of education and medical facilities are the internal impetus to the flow of urban-to-urban floating population, while comprehensive consideration will be given to obstacles such as money costs and psychological pressure generated by space friction; at the same time, the existing urban-to-urban floating population will reduce the flow costs, and stimulate the flow of urban-to-urban floating population according to existing routes by reducing the cost of inflows, and providing living and employment information. According to the stage division of population flow in the developed countries, it is discovered that the population movement in China is still transforming from the stage where people move from small cities and towns to metropolises. At this time point, the government should encourage the flow of population to large cities according to the natural law of population movement, decentralize the functions of the central area through the development of sub-centers and satellite cities, and promote the development of urban suburbanization, which are in line with the natural logic of urban development.

China is being at the stage of rapid urbanization. Based on the fifth census data, this paper analyzes the migration structure of China in different administrative and scale cities in the 1990s. The result shows that it is the prefecture cities that plays important role in the Chinese population migration in the 1990s, in which the big cities, especially the mega-cities, take a more significant position. It is also found that there are differences in the migration scale between county-level cities. The average migration scale of county-level city in East China is more than in West and Middle China. For the prefecture and above level cities is more important in the process of China′s urbanization at present and in the future, this article further examines the patial distribution pattern of inner-provincial and inter-provincial migration. It is showed that there are three ladder-like levels for prefecture cities in the spatial distribution pattern of inter-provincial migration: the first level includes the Zhujiang River Delta agglomeration, Changjiang River Delta city group, west bank cities, Beijing-Tianjin-Henan city agglomeration, the core cities of Liaodong Peninsula and a few of provincial capitals in middle-west area. The second level includes most of provincial capitals in middle-west area: Shandong Peninsula urban agglomeration and other eastern cities which are close to Beijing-Tianjin-Henan area, Changjiang River and Zhujiang River. The third level includes other provincial cities of middle west area and other cities which are far away from the core cities of three metropolises in the eastern area. Besides, the circle structure is obvious in the southeast coastal area, while the peak structure is more evident in the mid-west area. With regard to the spatial distribution pattern of inner-provincial migration, it shows that there are great similarities among prefecture-level in population spatial distribution pattern. The first gathering cities are each provincial capital cities. The ESDA technology is used to study the spatial distribution features of migration in cities in China. The result shows strong concentration for spatial distribution. The Zhujiang River Delta agglomeration, Changjiang River Delta city group are the main gather cities in both of inner and inter provincial migration. Furthermore, the migration gathering city range in Zhujiang River Delta area is larger than the Changjiang River Delta. On the contrary, the cities with small migration scales are located in agricultural provinces such as Anhui, Henan, and Sichuan, etc.

In recent years, urbanization has been attached an increasing importance in China's overall development strategies. Migrants' settlement intention in cities has played an important role in affecting the urbanization trend in China. In such a context, both scholars and policy makers have increasingly attempted to understand the settlement intention of migrants in China. However, few studies so far have examined the role of migrants' access to urban public services in affecting their settlement intention from the perspective of different-sized cities. Based on the data from "the 2016 national dynamic monitoring survey of migrant population" in Shanghai Municipality and Shandong, Jiangsu, Zhejiang, Fujian, and Guangdong provinces, this paper aims to address this issue. We utilize a composite index consisting of three dimensions to measure migrants' settlement intention, namely the long-term residence intention, the hukou transfer intention, and the urban housing purchase intention. We divide urban public services into two types, namely employment-related public services and social (non-employment) public services. The paper then explores the differences in the supply of the two types of urban public services and their impacts on migrants' urban settlement intention of different-sized cities. The results show that the bigger the size of a city is, the greater the likelihood that migrants get access to urban pubic services and the higher their level of urban settlement intention is. Migrants with easier access to the urban public services are more likely to settle down in cities. Through the ordered logistic regression model analysis, we also find that, after controlling for the effects of individual characteristics, family features, and migration characteristics, the two types of urban public services provided by cities for migrants have played a critical role in affecting migrants' urban settlement intention. However, the impacts of access to urban public services on migrants' urban settlement intention are different among different-sized cities. This is reflected in the following facts. First, migrants with easier access to unemployment insurance, medical insurance and housing security are more likely to settle down in cities, but this positive effect is limited to large cities. Second, in all size cities, migrants who are more likely to be covered by resident health records and receive more health education are more likely to settle down in cities. Finally, in small cities, only those covered by resident health records and receiving more health education have great effects on urban settlement intention.

It is demonstrated that the determinants of China's urban floating population's settlement intention are different among geographic units, which seems to be ignored by previous researches. Based on the data from the 2015 national migrant population dynamic monitoring survey (CMDS) and related statistics, this article uses the Semiparametric Geographically Weighted Regression (SGWR) model and k-means cluster method to examine the spatial variation of the factors influencing floating population's settlement intention in 282 prefecture- and provincial- level cites of China. Results provide the following conclusions. (1) The settlement intention of urban floating population is mainly influenced by the floating population characteristics instead of the destination characteristics. (2) Social and economic factors are closely related to the floating population's settlement intention. Meanwhile, the demographic, family and mobility factors exert a significant impact on such an intention. To be specific, there exists an inhibitory effect on floating population's settlement intention in factors such as income, marriage, and cross provincial mobility. However, housing expenditure, participation rate, number of children and other factors can effectively contribute to such intention. (3) Zonal spatial differentiation patterns of the influencing factors' coefficients are illustrated by the SGWR model, which can be further divided into four categories ("E-W", "N-S", "NE-SW" and "SE-NW"): The positive influences of ethnic and family factors are decreasing from the northern to southern regions, while the influence of employment ratio in the secondary industry is declining from the northwest to the southeast regions, and the impacts of factors such as the number of children and per capita GDP are diminishing from the northeast to the southwest regions. In eastern developed areas, the settlement intention of floating population with higher income is comparatively lower, while migrants with higher housing expenditure in southern China have a stronger intention to settle down. (4) Four influencing zones are detected by the k-means method: Floating population's settlement intention in North China, Central China and East China is significantly affected by multiple factors; In the northwest region and part of the southwest region, migrants' settlement intention is mainly influenced by demographic and social factors; The northeast region and the eastern part of Inner Mongolia's floating population's willingness to stay is mainly related to economic and family factors; Apart from housing expenditure, coefficients of other factors are relatively small in southern China and part of the central, eastern and southwestern regions. Additionally, this paper puts forward some suggestions on the service and management of the floating population in China.

China's rapid urbanization and economic development have given rise to the fast growth of the floating population, and housing is a key issue in the process of their integration into the destination cities. This paper intends to explore this topic by analyzing the spatial patterns of housing conditions of the floating population. Based on the sixth census data, the paper selects six indicators to measure housing conditions of the floating population: the home-ownership rate, the rental-housing rate, the floor area index, the housing facilities index (constructed by summing up the situation of five variables: availability of running water, washroom, bathroom, kitchen, and the type of fuel), the index of privacy (constructed by summing up the situation of two variables: the function of the dwelling and the number of the dwelling's floors), and the housing consumption index. It uses the methods of Spatial Autocorrelation Analysis and Hierarchical Cluster to examine the spatial distribution and agglomeration patterns of the floating population's housing conditions. The results of the calculation show that compared with urban permanent residents, members of the floating population are much more likely to live in rental homes;their housing conditions are generally worse;and their rental expenses are higher. The spatial variation of the homeownership rate, the rental-housing rate, and the housing facilities index is mainly manifested as north-south differences;the floor area index, and the index of privacy show marked difference between eastern and western China. The low-value centers of the housing consumption index are located in Inner Mongolia, Shaanxi, Hubei and Anhui provinces, while the high-value centers are located in Beijing. Furthermore the results of Spatial Autocorrelation Analysis demonstrate that there is a significant positive spatial correlation in the indicators of the floating population's housing conditions on a national scale, and identify the phenomenon of their spatial clustering and the centers of such spatial clustering. The analysis of Hierarchical Clustering identifies the housing conditions of the floating population into four distinctive groups, and suggests that the housing conditions of the floating population in the inner and east parts of China are better than those in the outer and west parts, and such a spatial variation extends from the north to the south. Finally, on the basis of the above findings, the paper puts forward some policy suggestions for improving the housing conditions of the floating population.

China’s floating population is the biggest mobile population in the world. Most of its members are not able to settle down and be fully integrated into the destination cities, need to circulate between the places of origin and destination, and adopt multi-locational livelihood strategies. The bi-local or even multi-local housing arrangement is an important part of the floating population’s multi-locational livelihood strategies, and a growing body of literature has been devoted to understanding their housing conditions and their determinants in the destination cities in such a context. Nevertheless, within this growing body of literature, far less attention has been given to understanding the housing investment and its determinants of the floating population in the places of origin. Using data from a survey of the floating population in Fujian Province in 2015, this article intends to fill this gap by examining two forms of the floating population’s housing investment in their places of origin, namely village home renovation and housing purchase. The article estimates two binary logistic regression models to examine the underlying factors affecting the decision and location selection of the floating population’s housing investment in their places of origin, and the main conclusions are as follows: 1) Housing investment in the places of origin is quite common among members of the floating population, with village home renovation being more widespread than housing purchase as the form of housing investment. 2) The floating population’s housing investment at their places of origin has been significantly affected by the periods of the investment and the life course of the investors, resulting from the interaction among China's macroeconomic situation, relevant government policies, and life events of individual members of the floating population. 3) Life courses of individual members of the floating population, their multi-locational livelihoods strategies, and community characteristics of their places of origin jointly determine both the decision and location selection of the floating population’s housing investment in their places of origin. The individual life course is the direct force determining floating population’s housing investment in their places of origin, and the community environment in the places of origin is the external driving force for such investment. The average annual amount of remittances and the times of returning home of the floating population play important and positive roles in promoting the housing investment of floating population in the places of origin, but the familization of the floating population’s migration and participation in the transfer of land use rights have a negative effect on the floating population’s housing investment in their places of origin.

Nowadays family migration has become the main migration pattern of floating population in China, which may raise new requirements for residential independence, tenure and quality in the cities. For some researchers, the migrant households were generally considered as passive recipients of housing inequality, and the institutional barrier of Hukou were no doubt still the decisive factors for migrants' housing outcomes. However, other researchers start to challenge this perspective, and argue that migrants should be considered as enabling agents with coping strategies in the housing market since the influence of Hukou system in migrants' housing outcomes is declining. Under this context, this paper tries to explore the role of family strategy in family migrants' housing outcomes, and whether this conclusion differentiates among various households. Based on a questionnaire survey which covered 2394 migrants and their households in 12 cities of the Bohai Rim Region, the Yangtze River Delta Region, the Pearl River Delta Region, and the Chengdu-Chongqing Region, this paper uses logistic regression model to explore family migrants' urban housing outcome in terms of housing tenure and quality and the mechanism for this. The result shows, different household arrangements lead to diversification of housing outcome, which means that the ownership and housing quality differs significantly among sole migrants, couple migrants, two-generation migrants, and three-generation migrants. The three-generation migrant households, as expected, have the highest ownership percentage and residential quality among all types of migrant households, while the sole and couple migrants suffer from the poor residential conditions. This housing outcome should be explained by factors such as socio-economic characteristics, household arrangements, geographic environment, migrants' connection with the original hometown and adaptation to current destination. Especially, when considering the effect of migrants' sense of belonging, plan to settle down in the destination and connection with hometown, the effect of Hukou is weakened. Different household arrangements have different barriers and strategies, which leads to the divergence of housing outcomes. The implication from this research is that the governments should make gradient regulation policy for the diverse family migrants, and lead them to settle stably step by step.

Based on data from migration dynamic micro survey 2016, this article first construct the network of interprovincial Hukou transfer intention in China. Then, with the use of some spatial analysis methods, including global Moran’s I coefficient and hot spot, we explore the spatial pattern of such network. Considering network autocorrelation in our data, Eigenvector Spatial Filtering Gravity Model (ESFGM) is applied for analyzing the driving factors. Main results reveal that: 1) The network shows a concentrating spatial pattern, and migrating flows with high ranks are mainly from the undeveloped areas to developed regions. That is to say, the willingness of transferring Hukou differs from regions to regions, showing heterogeneity. 2) Although there exists a random spatial pattern of Hukou attractiveness, such pattern of out-migration is concentrating. Furthermore, provinces in the west and northwest are the hot spot areas of Hukou emigration intention, while the cold spot areas are the mid-east regions in China. 3) From the macro perspective, population of destination shows a negative impact on Hukou transfer intention, while the size of population in an origin is not highly correlated; per capita GDP of both origin and destination, as well as export of foreign-invested firms, influence the intention positively and notably. But among all macro factors, the average wage of employees in urban areas leads to relatively higher impact. 4) From the micro perspective, generally speaking, migrants’ individual and family factors have strong ties with their Hukou transfer intention. A migrant, with higher education level, smaller age and larger scale of family in the destination, tends to transfer his (or her) Hukou to the immigration place. In the meantime, impacts of house condition and migration reason cannot be ignored: a migrant with lower intention of buying a local house or migrating for business, is likely to have a lower intention of transferring Hukou.