Local Government’s Land Finance Dependence and Migrants’ Settlement Intentions: Evidence from China

Abstract

In the context of China’s new urbanization construction, it is crucial to deal with the issue of migrants. While land finance, as an essential means of local government financing in China over the past three decades, has made important contributions to China’s economic growth, the impact of land finance on the settlement intentions of migrants still needs to be studied. This study considered cross-sectional data containing 78,642 samples based on the 2017 China Migrants Dynamic Survey 2017 (CMDS) and China Land and Resources Statistical Yearbook. In addition, a binary choice model was constructed, and the impact of the degree of land finance on the settlement intentions of migrants for the prefecture-level cities in China was empirically analyzed. With the help of the mediating- and moderating-effect models, the impact of the degree of land finance on the settlement intentions of migrants was examined from the perspectives of housing pressure, social security, and housing property rights. The results show the following: (1) The higher the degree of land finance, the lower the settlement intentions of migrants. (2) Land finance’s effect on settlement intentions is stronger for urban domiciles, low-skilled laborers, in large cities, and eastern and southern cities in these samples. (3) The higher the degree of land finance, the greater the housing pressure on the migrants, further reducing the settlement intentions of these individuals. However, social security and homeownership can reduce the negative effect of land finance on the settlement intentions of migrants. This paper provides policy implications for the construction of new urbanization and the realization of modernization.

1. Introduction

According to the Seventh National Population Census data released by China’s National Bureau of Statistics in 2021, the total number of migrants in China was 375 million in 2020, representing a significant but weak increase in the number of migrants compared to 2010. It reflects the general trend in a gradual slowdown in the growth rate of urbanization in China and the urgency of new urbanization. Therefore, the 14th Five-Year Plan for the Implementation of New Urbanization proposes to promote a new urbanization strategy with people as the core, to promote the civilization of the migrants continuously, and to improve the urbanization pattern with urban clusters as the main form and the coordinated development of large, medium and small cities and towns.

Migrants’ settlement intentions refer to the migrants’ intention to be permanently in the inflow city and not to return to their original residence. It is an indicator of whether the migrants are willing to stay permanently in the inflow city, which reflects not only the tendency of individuals to stay permanently in the inflow city at a micro level but also the attractiveness of the inflow of city at a macro level as well as the migration pattern and urbanization of the migrants in the region and even the whole country [1]. Therefore, it is of great practical significance to research the settlement intentions of migrants for China to build a new type of urbanization and to provide some references for other developing countries. However, local governments need economic strength and fiscal revenue to make migrants want to stay in an area. Since China’s 1994 tax reform, most of the local government’s revenue has come from land finance: the sale of land to generate revenue. However, at the same time, the policy of land finance may push up housing prices, causing certain social and economic impacts, and the impact of this impact on the migrants’ settlement intentions in the city has been less discussed, which is the gap this study is trying to fill.

Why do people choose to migrate? In fact, population migration is a complex issue. Any study that takes one factor as the main cause needs to be revised, and various disciplines have given different theoretical explanations. These theories can be divided into two categories, macro and micro theories. The macro theories mainly include Lewis’ dual economic theory [2], the laws of migration by Ravenstein [3], and push-pull theory proposed by Lee [4]. The micro theories mainly include the human capital theory and the new economics of migration theory. The research on the settlement intentions of the mobile population is abundant, and researchers have analyzed the causes and mechanisms of migration from economic, social, human capital, residential, and macro policies [5]. Among them, economic factors have been considered by many researchers as the most important factors [6,7], but some researchers have put forward different views. Firstly, economic factors are not always more important than others; some high-income earners in the U.S. do not have settlement intentions [8]. Secondly, the impact of economic factors (such as income growth) on the settlement intentions of migrants is not linear but inverted U-shaped after reaching a certain threshold and will inhibit the settlement intentions of migrants [9]. In terms of social factors and human capital factors, social identity [9], social ties [10], social integration [11], and even vocational-skills training [12] all have the significant effects on the settlement intentions of migrants. Finally, some residence factors, like housing security or homeownership, also have a significant effect on the settlement intentions of migrants [13,14].

Land finance is a policy that local governments in China adopt to increase fiscal revenue. It includes land tax, land transfer income, land rent, and business tax for real estate and construction, with land transfer income accounting for the largest share. Because land ownership in China belongs to the government; any individual, company, or group can only own the right to use the land; land transfer only includes the right to use, not ownership. It is a special fiscal system in China that has been seen as a powerful contributor to China’s economic growth in the past years [15] and was also the variable of interest in this study. Many studies on land finance have focused on urban research, and researchers have found that land finance has a positive impact on urban economic development. However, the intensity of the impact is related to the size of the city, and not all cities are affected equally: some cities may have a negative impact [16]. In addition to the economic impact, land finance significantly increases the speed of urban expansion, and cities under greater financial pressure have been more affected. Another important impact of land finance is on housing prices. A study showed very rapid and significant increases in housing prices in China [17]. The rapid rise in housing prices will have a negative impact on society. First, rising housing prices are spatially contagious, affecting neighboring provinces or cities [18]. Then, a study suggested that the rising housing prices are a generative impetus for the frequent occurrence of criminal offenses [19]. Rising housing prices can also lead to a labor “crowding-out” effect, inhibiting urban diversification and weakening the effect of industrial structure upgrading [20].

We find that there seems to be a lack of discussion of the relationship between the two in the literature on settlement intentions and on land finance. In fact, land finance, as an important policy implemented in China, has an all-embracing effect on urban residents, and, therefore, this study was dedicated to remedying the shortcomings of past studies on this issue. The most discussed macro policy in the early years was household registration [21], where researchers believed that the absence of an incoming household registration was the most important, if not the only, reason why migrants failed to settle in the incoming area. However, this view is problematic because migration is common in some countries where there is no household registration [22,23]. In recent years, other aspects of macro policies have attracted the attention of researchers, such as public service or social security, all of which have an impact on the settlement intentions of migrants [24,25]. The increase in public service is conducive to the increase in the settlement intentions of migrants, and this increase more significantly impacts the settlement intentions of women, less-educated, and rural migrants [26]. It is worth mentioning that some researchers are keen to identify the apparent negative effect of housing prices on migrants’ intention to settle [27], but this is not the essential reason because, for the government, the increase in housing prices is just a chain reaction after the adoption of land finance, so further research is needed in this regard.

Land finance is a policy factor that has driven China’s rapid economic growth over the past decades and was taken into account in our study. There needs to be more discussion on land finance. Compared with previous studies, the possible marginal contributions of this study are as follows: First, based on the perspective of land finance, this study enriches the understanding of the influence on the settlement intentions of migrants. Second, through theoretical analysis and empirical tests, the study attempted to determine the influence of land finance on the settlement intentions of migrants. Third, through theoretical analysis and empirical tests, we attempted to explore the heterogeneous characteristics of land finance on the settlement intentions of migrants and its specific mechanisms of action to help the academic community enrich the corresponding theoretical foundation.

2. Theoretical Analysis and Research Hypotheses

In this section, we analyze the research questions of this paper using four hypotheses. First, Hypothesis 1 corresponds to the first research question of this study: Does land finance affect migrants’ settlement intentions, and if so, does it have a positive or negative effect? Second, Hypotheses 2–4 correspond to the second research question of this study: What is the mechanism of action of land finance on migrants’ settlement intentions? Through which variables does it affect settlement intentions?

Since the 1994 tax reform, China’s local governments have received significantly lower taxes from the central government; at the same time, the central government also put a piece of land revenue, which at that time was still insignificant, into the hands of local governments, and no one seemed to have thought that this move would make local governments’ revenue increase instead of decrease. Local governments have used this opportunity to rapidly expand cities and promote infrastructure development. China is moving forward at speed unprecedented in human history. China has been in a period of rapid urbanization. Since the reform and opening up, a large number of people have flowed from the countryside to the cities, and data show that by the end of 1991, China’s urban population had increased to 312.03 million, an increase of 80.9% over 1978, with an average annual growth of 5.8%. The urbanization rate reached 26.94%, 9% higher than in 1978. At the same time, the history of the economic rise of Western countries shows that the less-efficient tax finance could not fully meet the demand for primitive capital in the start-up phase of urbanization. Therefore, after combining the experience of Western countries with Chinese characteristics, China chose to complete the original accumulation of capital in the initial stage of urbanization through land finance, rapidly expanding infrastructure investment, attracting foreign investment, changing the urban landscape, and promoting the development of the local economy [15].

In fact, the positive role of land finance as a means of financing for local governments in the early days should be fully recognized. Through fiscal decentralization, the central government triggered healthy competition among local governments to build land, boosting infrastructure construction and economic growth, and most of the country’s population enjoyed the benefits of development. However, as China’s economy and urbanization have slowed, local governments have not found other ways to expand their fiscal revenues significantly, so they continue to increase the area of land transfer and increase their reliance on land finance, which poses serious problems. First, land finance creates houses as investment properties. Buying a city’s real estate is equivalent to buying the city’s stock and the city’s economic growth. This is followed by a rapid rise in housing prices; as the government has no way to suppress housing prices through the logic and means of the general commodity market and capital, capital will madly rush into the real estate market, attracting speculators and other industries, especially the manufacturing industry. This will have a great impact and cause financing difficulties; only those industries related to real estate can achieve growth, thereby further squeezing the employment opportunities of people in other industries and reducing the income of the migrants and social integration. This, in turn, reduces migrants’ settlement intentions [28,29]. Second, after the financialization of real estate, due to its high entry barrier, the wealth gap between those who do not acquire property and those who own property will widen more and more. With the rapid rise in housing prices, those who have a home can rapidly increase their assets even if they do not work hard, while those who do not have a home can hardly keep up with the growth of housing prices even if they work hard, so that the poverty gap will increase with the rise in housing prices [30]. This poverty gap in a city makes the migrants not see the possibility of class mobility, which affects the social and economic status and social mentality of the migrants and reduces the migrants’ settlement intentions. Third, land finance affects the financial capacity of local governments, reduces the willingness and motivation of governments to invest in other industries, and causes a waste of resources. Local governments continue to invest a large amount of resources that should be used for other industries into real estate and blindly expand cities. On the one hand, these behaviors hinder the upgrading of the industrial structure and regional technological innovation, and, on the other hand, urban expansion leads to air pollution and environmental destruction, which damages the health of migrants and ultimately reduces their willingness to settle [31]. Based on the above analysis, this paper proposes research Hypothesis 1.

Hypothesis 1 (H1).

Land finance can reduce the settlement intentions of migrants.

The most intuitive and biggest impact of land finance on migrants is the rapid rise of housing prices, and it is difficult for the income growth of the migrants to keep up with the speed of property price growth. Under the current tax system in China, the maximum marginal tax rate of 45% is applied to labor, while the tax rate for capital is fixed at 20%. We find that the labor tax is obviously higher than the capital gains tax. Under such circumstances, migrants are increasingly unable to afford local housing, and China’s household registration system determines a huge gap between those with and without housing in accessing public services such as education, medical care, and cultural products provided by local governments. The long-standing gap has created housing pressure. According to the theory of political economy, housing is a necessity for household labor reproduction, and the housing pressure caused by rising housing prices affects the labor reproduction of households, which then affects the income and consumption of laborers [32]. In addition, most of the migrants take out loans to buy houses, and the mortgage accounts for a high proportion of their income. This housing pressure places pressure on the rest of migrants’ consumption and savings and, at the same time, makes them increase their work hours in exchange for labor compensation, which reduces their job stability and psychological health [33]. This finally leads to a decrease in the migrants’ settlement intentions. Based on the above analysis, this paper proposes research Hypothesis 2.

Hypothesis 2 (H2).

Land finance can reduce settlement intentions by increasing the housing pressure on migrants.

Social security, as a form of redistribution provided by the state or government to residents through laws or institutions, can effectively narrow the poverty gap and change the distribution of wealth. It is also an important means for China to move toward common prosperity [34]. This narrowing of the gap is not only reflected between high- and low-income earners but also between generations of young and middle-aged people or older people, between the east and the middle or west, and between individuals over their lifespan. Narrowing the income gap among the groups can significantly improve the subjective socioeconomic status of migrants, improve subjective well-being [35], and thus enhance subjective well-being and perception of social equity. In addition to income, social security can also improve human capital investment or sustainable economic growth [32,36]. According to social support theory, a formal social support system can improve people’s ability to cope with difficulties and alleviate negative pressure. The stronger the social support network of a person, the better they will be able to cope with challenges from the environment. So, social security can play a hedging role and improve settlement intentions in the face of the negative effect played by land finance on migrants. Based on the above analysis, this paper proposes research Hypothesis 3.

Hypothesis 3 (H3).

Social security plays a positive moderating role in the impact of land finance on migrants’ settlement intentions.

According to housing class theory, different groups within cities can be divided into six categories based on differences in access to housing. These six categories have significant differences in their ability to use markets and policies (economic income), which also reflects the inequitable distribution of resources [37]. The literature also shows that the difference between the housed and the unhoused is significantly reflected in the settlement intentions and that this difference affects the social integration of migrants [38,39], income inequality, and subjective well-being [40,41]. For those who own a home, the increase in housing prices due to land finance has a “wealth effect” on them, increasing their rental income, financial support, and thus their consumption power. For those who do not own a home, the increase in housing prices caused by land finance has a “house poor effect” on them, forcing them to reduce their expenses, increase their working hours, and reduce their consumption opportunities in order to make a down payment. The double effect of income and consumption leads to a significant difference in the settlement intentions between those who own a home and those who do not. Therefore, homeownership can act as a hedge against the negative effect of land finance on migrants and increase settlement intentions. Based on the above analysis, this paper proposes research Hypothesis 4.

Hypothesis 4 (H4).

Homeownership can moderate the negative effect of land finance on the migrants’ settlement intentions.

3. Data and Methods

3.1. Data Sources

In this study, two data sources were used: The first was the China Migrants Dynamic Survey (CMDS), organized and implemented by the National Health Commission. The survey targets migrants aged 15–59 years old who have lived in the inflow area for over a month and are not registered in their district (county or city). The survey contains data on migrants in 31 provinces, with a total sample size of 169,889. This is a national survey with wide data coverage and good national representativeness. The CMDS consists of a survey of individuals’ household members and income and expenses, employment, settlement intentions, health and public services, and social integration in five modules. On average, there are about ten questions in each module. According to the research needs of this study, the samples of the migrants that clearly expressed whether they had settlement intentions or not were selected. The second source was city-level data, and the land finance data were obtained from the ratio of land transfer to government budgetary revenue for the prefecture-level cities in the China Land and Resources Statistical Yearbook of the corresponding year. The China Land Resources Statistical Yearbook mainly covers the status of China’s land resources and the administrative management of these land resources. The scope of statistics included the national land, mineral, and marine resources the investigation and survey of land resources; the administrative management of land resources and mineral resources by natural resources administrative departments at the national, provincial, municipal, and county levels; the national administrative management of marine resources; scientific and technological research on land resources; and mapping and geographic information. After data matching and the elimination of missing values, the basic sample size in this study was 78,642.

3.2. Methods

3.2.1. Selection of Model Variables

The explained variable was settlement intentions. The settlement intentions measured in this study were long-term settlement intentions. Therefore, the more short-term settlement intentions were treated as no settlement intentions, referring to the study of Yu [42]. We used the question, “Do you intend to live in the area long-term (5 years)?”, with an answer of “yes” having a value of 1, and an answer of “no” having a value of 0.

The explanatory variable was land finance. Referring to the study of Hou [43], the ratio of land transfer revenue to public fiscal revenue was chosen as a way to measure the explanatory variables. The level of this variable was city.

The moderating variables were homeownership and social security. First, referring to the study of Huang [44], we used the question “Do you have your own house?”, with an answer of “yes” having a value of 1, and an answer of “no” having a value of 0. Then, because the social security card is an important item issued by the Chinese government, it enables the receiving of a pension or retirement benefits, medical payments, and identity verification, so we used the question, “Do you have a personal social security card?”, with an answer of “yes” having a value of 1, and an answer of “no” having a value of 0.

The mediating variable was housing pressure. We used the question, “Do you have housing pressure?” as a criterion and assigned a value of 1 to “yes” and 0 to “no”.

The control variables mainly included demographic variables (sex, age, square of age, nationality, marital status, and monthly household income) as well as mobility characteristics (range, time, and mode). Descriptive statistics of the variables are shown in

Table 1. Variable definition and descriptive statistics.

Variable	Definition and Measure	Mean	Std. Dev.
Settlement intentions	Do you intend to live in the area long-term (5 years)? Yes = 1, No = 0	0.670	0.470
Land finance	The ratio of land transfer revenue to public fiscal revenue (City)	0.700	0.443
Mobility range	What is the range of your mobility? Across province = 1, otherwise = 0	0.488	0.500
Mobility time	Length of flow into the local area (Year)	6.587	6.115
Mobility pattern	rural-to-urban mobility = 0, urban-to-urban mobility = 1	0.193	0.395
Age	2017-the year of birth	35.791	9.476
Sex	Male = 0, Female = 1	0.424	0.494
Nationality	Han Nationality = 0, other = 1	0.922	0.269
Marital status	Married = 1, unmarried = 0	0.856	0.351
Monthly household Income	Logarithm of monthly household income	8.795	0.600
Social Security	Have a social security card = 1, unclear and no = 0	0.583	0.493
Homeownership	Do you have your own house? Yes = 1 and no = 0	0.312	0.463
Housing pressure	Do you have housing pressure? Yes = 1 and no = 0	0.598	0.490

.

3.2.2. The Models

Based on the available literature, the binary Probit model was constructed in conjunction with the purpose of this study as follows:

$$Y_{ij}={\beta }_0+{\beta }_1{LF}_{ij}+{\beta }_2{X}_{ij}+{Province}_j+{\epsilon }_{ij}$$

(1)

In Formula (1), Y represents the settlement intentions of the migrants and is a binary variable, i represents different individuals, and j represents different cities. β₀ is an intercept term, and β₁ is the estimated coefficient of the probability of the effect of land finance on the settlement intentions of migrants. LF is the explanatory variable of land finance; X includes other control variables, including individual characteristics and mobility characteristics; Province is the province fixed effect, and ε is the error terms.

However, there may be a causal bidirectional relationship between land finance and migrants’ settlement intentions. So, in order to address this problem, we attempted to find instrumental variables. In this study, the length of tenure of officials was used as an instrumental variable. Specifically, the tenure of the mayor of the prefecture-level city as of 2017 was selected as a proxy variable. On the one hand, mayors with longer tenure are more likely to use land transfer revenue to show their performance as soon as possible and therefore are more likely to grant commercial and residential land. There is a correlation assumption between the length of tenure of officials and land finance. On the other hand, since the length of tenure of officials is not directly related to migrants’ settlement intentions, we argue that this variable also satisfies the exogenous assumption of the instrumental variable.

4. Results

4.1. Baseline Regression

Table 2. Probit regression results.

Variables	Settlement Intentions
Land finance	(1)	(2)	(3)
	−0.147 ***	−0.150 ***	−0.534 ***
	(0.011)	(0.012)	(0.046)
Mobility time		0.058 ***	0.056 ***
		(0.001)	(0.001)
Mobility range		−0.544 ***	−0.566 ***
		(0.011)	(0.012)
Monthly household income		0.381 ***	0.395 ***
		(0.009)	(0.010)
Nationality		0.047 **	0.024
		(0.021)	(0.022)
Sex		0.067 ***	0.065 ***
		(0.010)	(0.011)
Age		0.039 ***	0.038 ***
		(0.004)	(0.004)
Age squared		−0.001 ***	−0.001 ***
		(0.000)	(0.000)
Pseudo-R2	0.035	0.1351	0.2323
Provincial fixed effects	YES	YES	YES
Wald chi2			136.93 ***
AR chi2			138.53 ***
Observations	78,642	78,642	74,140

Note: *** p < 0.01, ** p < 0.05; robust standard errors are in parentheses.

shows the regression results of Formula (1). Column (2) controls for province fixed effects. Column (2) shows the inclusion of mobility and individual characteristics variables. Column (3) shows the results of the instrumental variables’ analysis for the tenure of officials. As seen from the regression results in Column (1), the effect of land finance on the migrants’ settlement intentions is negatively significant at the 1% statistical level. For every 1% increase in land finance, the probability that the migrants’ settlement intentions is decreased by 14.7%, which supports Hypothesis 1 of this study. To test the strength of the instrumental variables, IV regression estimation was used, and the results showed that the Wald Chi2 and AR Chi2 values were 136.93 and 138.53, significantly rejecting the original hypothesis. So, there was no problem of weak instrumental variables, and the instrumental variables were reasonably chosen. The estimation results in Column (3) show that the effect of land finance on the settlement intentions of migrants is negatively significant (p < 0.01), and the coefficient of the IV regression is larger than that of the baseline regression, which indicates that the negative effect of land finance on the settlement intentions of migrants may be underestimated to some extent if the endogeneity issue is not considered. From the results of each control variable in Column (3), the mobility time, mobility range, monthly household income, marital status, sex, and age all significantly affect the settlement intentions of migrants. Specifically, the duration of mobility variable has a significant positive effect on the settlement intentions of migrants, and the longer the duration of mobility, the higher the settlement intentions of migrants. This may be because, on the one hand, the social culture, customs, and even climate and food of the inflow cities are very different from those of the domicile, and it is difficult for the migrants to adapt. On the other hand, interprovincial mobility means that most or even all of the existing social networks of the migrants are invalid, and they have to re-establish their social networks in the inflow cities, which easily creates a sense of “relative deprivation” and reduces the degree of social integration, thus affecting their settlement intentions. Monthly household income has a significant positive effect on the settlement intentions of migrants because the migrants with higher monthly household income are more likely to have sufficient social resources, access to urban public services, and a higher self-perception of class, which leads to a sense of “relative satisfaction” and thus increases the settlement intentions. The marital status and sex variables have a significant positive effect on the settlement intentions, indicating that the settlement intentions of women and married migrants are higher than those of men and unmarried migrants, probably because married migrants are more influenced by family factors than unmarried migrants, and the implicit cost of changing cities is higher, while the pressure of life in cities due to work and housing is often higher for men than for women, so the settlement intentions decrease. Men tend to have higher life pressure in cities due to work and housing than women, and their settlement intentions decrease. The positive effect of age on the settlement intentions of migrants and the negative effect of its squared term indicate that age has an inverted U-shaped relationship with the settlement intentions of migrants, and the settlement intentions of migrants increase and then decrease with age.

4.2. Robustness Test

This paper tested the robustness in the following aspects: First, we replaced the measurement of the explained variable. We used the question, “If you meet the conditions for local settlement, are you willing to move your household registration to the local area?” Household registration is an important system in China’s urban management, and those willing to move their household registration to the inflow area have a high degree of long-term settlement intentions. Second, we replaced the measurement of the explanatory variable. In China, public fiscal revenues currently exceeds expenditures in developed regions, and public fiscal expenditures exceed revenues in less developed regions; therefore, it was also feasible to use the ratio of land transfer revenue in public fiscal expenditures as a proxy variable for land finance. Third, we replaced the estimating model; the baseline model in this study was the Probit model; in this part, we replaced the Probit with the LMP model as well as the Logit model for regression. The results of the robustness test are shown in

Table 3. Robustness tests.

Variables	Settlement Intentions
	Replace the Explained Variable	Replace the Explanatory Variable	LMP	Logit
	(1)	(2)	(3)	(4)
Land finance	−0.153 ***	−0.246 ***	−0.052 ***	−0.254 ***
	(0.011)	(0.018)	(0.004)	(0.020)
Province Fixed	YES	YES	YES	YES
Control variables	YES	YES	YES	YES
Observations	78,642	78,642	78,642	78,642

Note: *** p < 0.01; robust standard errors are in parentheses.

. It can be seen that regardless of replacing the measurement of explained variable or explanatory variable or replacing the regression model, there is still a very significant negative effect of land finance on the settlement intentions of migrants, so the results of this study are robust.

4.3. Heterogeneity Test

4.3.1. Individual Heterogeneity

In order to understand the impact of land finance on different groups of migrants, this study conducted a heterogeneity test after controlling for other variables, and the results are shown in

Table 4. Probit regression results for different individuals.

Variables	Settlement Intentions
	Rural-to-Urban Migrants	Urban-to-Urban Migrants	Low-Skilled	High-Skilled
	(1)	(2)	(3)	(4)
Land finance	−0.122 ***	−0.204 ***	−0.142 ***	−0.068 **
	(0.013)	(0.032)	(0.013)	(0.030)
Province fixed	YES	YES	YES	YES
Control variables	YES	YES	YES	YES
Observations	63,462	15,178	46,425	32,208

Note: *** p < 0.01, ** p < 0.05; robust standard errors are in parentheses.

. The sample was divided into individuals with urban-to-urban mobility and individuals with rural-to-urban mobility, depending on their mobility patterns. From Columns (1) and (2) in Table 4, we can see that land finance has a significant negative effect on the migrants in both mobility patterns, but the effect is stronger for individuals in the urban-to-urban mobility pattern than in the rural-to-urban mobility pattern. The possible reason is that the rural-to-urban migrants move mainly for economic reasons, mostly to obtain higher income inflow cities, and a higher proportion of this income is returned to family members (such as the elderly and young children) who stay in the countryside. So, they have fewer savings, low willingness to purchase a house, low settlement intentions, only use the cities as a place to obtain income, and are less affected by the land fiscal policy. Comparing urban-to-urban with rural-to-urban migrants, urban-to-urban migrants have more savings, are more willing to buy houses, have a higher proportion of house purchase, and have higher settlement intentions in the city, and use the place of inflow more as a place to stay for a long time in the future. Therefore, the impact of land finance on urban-to-urban migrants is stronger than that of rural-to-urban migrants.

Second, based on educational heterogeneity, individuals with college and higher degrees are considered as high-skilled labor and the rest as low-skilled labor. From Columns (3) and (4) in Table 4, we can see that land finance has a significant negative effect on both types of labor, but the effect is stronger for low-skilled labor than for high-skilled labor. The possible reason is that land finance triggers an increase in housing prices, affecting local firms’ capital–labor ratio [45]. There is a substitution of capital for labor, and technological progress is more capital-biased [46], which has a further destructive effect on low-skilled labor, reducing their job opportunities and labor compensation and reducing the settlement intentions of low-skilled labor.

In contrast, for high-skilled labor, the capital-biased technological advancement brought by land finance creates more high-level tasks, and enterprises increase their demand for high-skilled labor. In contrast, high-skilled labor has better “bargaining power” in the labor market. The rising wage level can partially offset the negative impact of rising housing prices caused by land finance and reduce the negative effect of land finance on the settlement intentions of high-skilled labor. Thus, the impact of land finance on low-skilled labor is stronger than that on high-skilled labor.

4.3.2. Regional and Urban Size Heterogeneity

According to the 2021 China Population Migration Report, the current general trend of China’s population movement has always been to gather in the south and east, with the average annual resident population growth in the south being nearly six times that of the north in the past ten years, and the population share in the east rising by 2.01 percentage points. The second trend is to gather in a few core cities, with the average annual population growth rate in the first-tier cities in the past ten years being 2.37%. The average annual population growth rate in the second-tier cities has been 1.91%. In addition, within migration, rural migrants make up the majority of the population, and it is predicted that of the 1.3 billionnew urban population added over the next ten years, 0.7 billion will come from rural migrants. Therefore, based on this macro context, in addition to individual heterogeneity, regional heterogeneity may be reflected in the effect of land finance on the migrants’ settlement intentions. For this reason, China’s geographical regions were divided into east, middle, and west regions; meanwhile, due to the continuous southward shift of China’s economic center of gravity and the increasingly obvious difference between north and south, China was divided into south and north. Finally, the cities in the sample were divided into large cities (first tier and second tier), small cities (third tier and fourth tier). After controlling for other variables, the specific regression results are shown in

Table 5. Probit regression results for different regions and city sizes.

Variables	Settlement Intentions
	Region					City Size
	South	North	East	Middle	West	Large	Small
	(1)	(2)	(3)	(4)	(6)	(7)	(8)
Land finance	−0.166 ***	−0.038	−0.184 ***	0.013	0.194 ***	−0.219 ***	−0.005
	(0.014)	(0.027)	(0.015)	(0.028)	(0.038)	(0.016)	(0.020)
Province fixed	YES	YES	YES	YES	YES	YES	YES
Control variables	YES	YES	YES	YES	YES	YES	YES
Observations	46,425	32,208	39,612	17,037	21,979	48,809	29,822

Note: *** p < 0.01; robust standard errors are in parentheses.

.

From Columns (1) and (2) in Table 5, it can be seen that the dependence on land finance has a serious “crowding-out” effect on the migrants in the south of China, which reduces the migrants’ settlement intentions; at the same time, although there is a certain negative effect on the north, the effect is not significant. The possible reason is that the economic systems in the north and the south are very different: the reform of the economic system mechanism in the north is seriously lagging behind that in the south, and the economic development in the north relies heavily on large state-owned enterprises. For example, in 2017, the proportion of the main business income of non-state-controlled enterprises in the north to industrial enterprises above the scale was 58.4%, while the proportion in the south was 78.9%. The level of government intervention in the north is higher than that in the south, but the efficiency of resource allocation is relatively lower. The level of capital market development is low, and the economic vitality in the north is significantly lower than that in the south, which leads to the desire of local governments in the north to attract investment and grant land less than in the south. The motivation of local officials in the north to develop the economy is lower than that in the south, so the impact of land finance is stronger in the south than in the north.

From Columns (3)–(5) in Table 5, it can be seen that the “crowding-out “ effect of land finance on the migrants is only significant in the eastern region, where land finance significantly reduces the settlement intentions of migrants. For the middle and western regions, it has a “spillover “effect. The “spillover” effect is even higher in the western region than in the eastern region. The possible reason is that the effect of land finance on the settlement intentions of migrants is influenced by economic growth, and the relationship between land finance and economic growth is nonlinear, with an overall inverted “U” shape. In the early stage, land finance can increase the fiscal revenue of local governments so that local governments can use the revenue from land transfer to build infrastructure and lower the tax revenue of enterprises to attract investment and promote economic growth, while in the later stage, the proportion of the industrial sector in the regional economy decreases, the elasticity of infrastructure output may decrease, and the proportion of the service sector increases. The cost of living of residents increases, and their willingness to consume decreases, further hindering economic growth, which is confirmed by existing studies [47]. In other words, the eastern, middle, and western regions are in different periods when land finance affects economic growth, the Eastern region is in the later period when land finance hinders economic growth, while the Middle and Western regions are in the first and Middle period when land finance drives economic growth, and the income increase brought by economic growth hedges the effect of land finance on the migrants’ settlement intentions. Therefore, land finance has a “crowding-out” effect on the migrants in the eastern region and a “spillover“ effect on the migrants in the middle and western regions.

From Columns (6) and (7) in Table 5, we can see that land finance has a serious “crowding-out” effect on the migrants in large cities, significantly reducing the migrants’ settlement intentions, while it has a negative but not significant effect on the migrants in small cities. The possible reason for this is that the Chinese government is trying to control the influx of people from big cities. So, compared with small cities, large cities have a stricter household registration system and have certain entry thresholds, while land finance continues to push up housing prices in large cities. The high housing prices makes it unaffordable for most migrants to acquire household registration in large cities through homeownership and the education, medical care, technology, infrastructure, and other public services that come with household registration. The lack of these public services creates a sense of relative deprivation when compared with the local population in large cities, which in turn reduces the degree of social integration and affects the settlement intentions. For small cities, although the dependence on land finance also drives property prices in the small cities, compared with large cities, the housing prices are still within the affordability of the migrants, and the management of household registration is more relaxed, so the migrants can easily obtain household registration and the corresponding public services. Additionally, the possibility of “relative deprivation” decreases compared with that of the locals as a reference group, and the degree of social integration is higher compared with that in large cities. They have a higher degree of social integration and higher settlement intentions than those at large. Therefore, the dependence on land finance has a “crowding-out” effect on the migrants in large cities, while it has some negative but not significant effects on those in small cities.

4.4. Mechanism Analysis

While land finance enriches local government revenues, it also contributes, to a large extent, to the increasingly higher housing prices. The growth in residents’ income cannot keep up with the increases in housing prices, and the proportion of housing expenditure to residents’ expenditure keeps increasing, which in turn causes housing pressure. So, does this pressure transmit to the settlement intentions? Therefore, we conducted a test of the mediating effects in this section. The results are shown in Columns (2) and (3) in

Table 6. Mediation and moderating effect results.

Variable	Housing Pressure	Settlement Intentions
	(1)	(2)	(3)	(4)
Land finance	0.029 ***	−0.004 ***	−0.150 ***	−0.154 ***
	(0.005)	(0.025)	(0.012)	(0.012)
Housing pressure		−0.106 ***
		(0.004)
Land finance × social security			−0.048 **
			(0.022)
Land finance × homeownership				−0.042 **
				(0.023)
Province fixed	YES	YES	YES	YES
Control variables	YES	YES	YES	YES
Observations	46,425	32,208	39,612	17,037

Note: *** p < 0.01, ** p < 0.05; robust standard errors are in parentheses.

, which show that land finance can significantly increase housing pressure, and the housing pressure has a significant negative effect on the settlement intentions of migrants (p < 0.01), which indicates that housing pressure significantly reduces the settlement intentions of migrants. The land finance of local governments is transmitted to the housing pressure of the migrants, which in turn affects the settlement intentions of migrants. These findings validate research Hypothesis 2 in this study.

Social security, as a form of wealth redistribution and the main body of the fiscal redistribution effect, can play the role of income transfer, promote consumption, and alleviate the poverty gap [37]. According to the data, the largest proportion of assets of Chinese residents is property; then, can social security play a role in regulating the series of negative effects triggered by the dependence on land finance? To this end, this paper empirically tested the moderating role of social security, and the results are presented in Column (4) in Table 6. The empirical results show that the coefficient of land finance on the settlement intentions of migrants is −0.15, which passes the significance test at the 1% level, while the coefficient of the interaction term between land finance and social security is −0.048, which passes the significance test at the 5% level. This indicates that the negative effect of land finance on the settlement intentions of migrants is negatively regulated by social security, and having social security can significantly reduce the “crowding-out” effect of land finance on the settlement intentions of migrants, which supports Hypothesis 3 in this study. Therefore, local governments should guide the incoming population to actively participate in social security, establish a sound multilevel social security system, expand the social security fund pool, narrow the gap between urban and rural areas and between mobile and non-migrants, and establish a more equitable social security system to reduce the negative effects of land finance.

For migrants, owning a home means having a household registration in the place of inflow, which means having public services such as education, medical care, social security, cultural products, and basic social services provided by the local government. Those who own their own homes have a higher degree of social integration or class identity than those who do not have a home [42]. Then, can housing property rights play a role in regulating the negative impact of land finance on the migrants’ settlement intentions? To this end, we empirically examined the moderating effect of housing property rights, and the results are presented in Column (5) in Table 6. The results show that the coefficient of land finance on the settlement intentions of migrants is −0.154, which passes the significance test at the 1% level, and the coefficient of the interaction term between land finance and housing property rights is 0.042, which also passes the significance test at the 5% level. This indicates that the negative effect of land finance on the settlement intentions of migrants is negatively moderated by homeownership, and owning a home can significantly reduce the “crowding-out” effect of land finance on the settlement intentions of migrants, which supports Hypothesis 4 in this paper. Therefore, local governments should promote the supply of subsidized housing and provide a large amount of subsidized housing for the migrants to reduce the negative effect of land finance, increase the migrants’ settlement intentions, and promote new urbanization.

5. Discussion

Land finance has greatly contributed to China’s rapid economic growth over the past decades. China’s local governments have received large amounts of revenue through this policy. Then, they built up a good urban infrastructure and public services such as education and health care, high-speed rail, airports, and seemingly perfect economic data, which officials were very satisfied with. However, this was not sustainable. China’s vast land area has led to a huge amount of land owned by local governments. China’s administrative system ensures that land ownership will always belong to them, so it can be said that local governments are in a monopoly position in land transactions. Therefore, even if they face unsustainable land finance policies, they will still choose to continue them. After all, officials have a term of office, and economic growth during their term of office, as long as it is good, will create the possibility of promotion.

Any policy can bring both benefits and harms, and the same is true of land finance, which has tended to be ignored in previous studies concentrating on economic growth and urban expansion brought about by land finance [48], even though some scholars recognized the harms but focused on measures of urban management. These points are partially correct, but economic growth is not a numbers game. The purpose of economic development is to make the lives of those of the nation better. We should focus on the impact of a policy not only on cities but also on individuals, especially the impacts on some disadvantaged groups. In China, the biggest disadvantaged group is these migrants. These migrants have a specific official term in China: the floating population. The biggest difference between them and natives is that they do not have a household registration, and because they do not, many urban public services are unavailable. For them, more research is necessary.

Moreover, in the cross-country comparison, we found some peculiarities specific to China. First, other countries lack this policy of land finance. In the United States, for example, the largest source of revenue for state governments is individual income tax, while in China, it is the revenue from land transfers. Second, in China, land ownership is in the hands of the government, which only sells the right to use it during land transactions, while in the U.S., both can be sold. The U.S. has property taxes, while China does not. So, we can see that China’s land transfer revenue is more like a hidden tax on the people. These characteristics mean that land finance’s negative impact on migrants’ settlement intentions may only occur in China. In other regions, such as the U.K., a study showed that high housing prices in potential destinations discourage migration and that an important determinant of migration is the cost of housing [49], which is similar to some of the points made in this study but did not involve a discussion of land finance. These specific qualities were also of interest to this study.

6. Conclusions and Policy Implications

We empirically investigated the effect of land finance on the settlement intentions of migrants based on data from the 2017 China Migrants Dynamic Survey (CMDS) and matched macro-level data from prefecture-level cities by building a Probit model, and the following conclusions were drawn: First, land finance significantly reduces the settlement intentions of migrants. This finding remained robust after using different estimation methods and considering the issue of endogeneity. Second, the effects of land finance on migrants’ settlement intentions differ across individuals or regions: the migrants in the rural-to-urban mobility pattern suffer fewer negative effects than the migrants in the urban-to-urban mobility pattern, and high-skilled laborers suffer fewer negative effects than low-skilled laborers. In terms of the south and north, land finance in the south has a significant in terms of north and south; the “crowding-out” effect of land finance on the migrants in the south and the “spillover” effect of land finance on the migrants in the north are significant. In terms of east, middle, and west, the “crowding-out” effect of land finance on the migrants was found. In terms of urban hierarchy, land finance has a serious “crowding-out” effect on the migrants in large cities, while the effect is not significant for small cities and below. Finally, housing pressure, social security, and housing property rights are important mechanisms through which land finance affects the settlement intentions of migrants. The negative effect of land finance on migrants’ settlement intentions is transmitted through the mediating variable of housing pressure. Social security and housing property rights, as moderating variables, can effectively mitigate the negative effect of land finance on the settlement intentions of migrants and reduce its influence.

Based on the above conclusions, the following suggestions are made in the context of China’s new urbanization construction: (1) On the national level, local governments should gradually reduce their reliance on land finance, adhere to the concept of “housing without speculation” and “people-oriented development”, protect the rights and interests of the migrants, narrow the gap between them and the household population, gradually remove the shackles of the household registration system on labor mobility, and promote equalization of basic public services. At the same time, the middle and local governments should speed up the construction of corresponding laws and regulations, improve the management of the real estate market, and gradually pull back real estate from the financial and capital markets to the general commodity market by means of “equal rights for rent and sale” and “rent first, sell later”, and vigorously combat speculators and house flippers. (2) Local governments should pay attention to the heterogeneous characteristics of the migrants and give priority to protecting their rights and interests. For low-skilled laborers and rural-to-urban migrant individuals, they should introduce corresponding policies to adjust the industrial structure and promote inclusive employment. As for each region, the eastern, southern, and large cities should take the lead and rapidly reduce their reliance on land finance, while the middle, western, northern, and third-tier cities and below should review the situation, respect the objective law of economic development, and appropriately adjust their land finance policies. (3) We need to vigorously promote the construction of subsidized housing, increase the number and coverage of subsidized housing, reduce the housing pressure of the migrants through subsidized housing, gradually reduce the share of the commercial housing market and have its share filled by subsidized housing, and reduce housing inequality. At the same time, we should gradually expand social security from broad coverage to full coverage, establish a sound social security system, improve the protection capacity, and promote the equalization of social security services through the construction of a unified national social security platform and the use of digital means to promote common prosperity. (4) Local governments should gradually move away the current land finance model, where land transfer revenue is the main source of public fiscal revenue, and find a new source of taxation by establishing a property tax-based taxation model that can adapt to China’s current economic development trend.