Effect of Land Marketization Level and Land Prices on Foreign Direct Investment in China

Abstract

With the increase in China’s land marketization level, the fundamental role of pricing mechanisms in optimizing the allocation of land resources has received greater attention and has affected foreign direct investment (FDI). This paper analyzes the effects of land marketization level and land prices on FDI using a mediation model based on data on primary land market transactions in 256 cities from 2003 to 2019. The results indicate that the land marketization level, land prices, and FDI are all increasing, with regional heterogeneity. Compared with low-level cities, the land marketization level and land prices in high-level cities have increased the most. Low-level cities have become new hotspots for FDI, whereas high-level cities remain the main destinations for FDI. Both land marketization levels and land prices significantly promote FDI. Land marketization level exerts a weak direct positive effect on FDI in high-level cities but a strong indirect positive effect on FDI in high-level cities through land prices in general. The results of this study enrich the literature on the location decisions of multinational corporations and may serve as a reference for China and other developing countries in developing policies to attract FDI.

1. Introduction

With the rapid pace of globalization in today’s world, foreign direct investment (FDI), mainly by multinational corporations (MNCs), is expanding rapidly worldwide and significantly affecting the economic development of developing countries. Given the improvement of related infrastructure and resources and the soundness of the legal and regulatory system in China, many MNCs have moved their production, research and development, and management departments to China to benefit from its relatively low production costs and vast market. In 2020, FDI in China reached $163 billion, and China surpassed the United States as the largest foreign investment destination in the world [1]. Although FDI inflow has increased regional capital investment and helped China obtain advanced technologies and experienced management talent, it has also exerted certain negative effects on urban development. For example, it has contributed to an increase in land prices, especially in the coastal region of China [2]. This increase in land prices has reduced the expected return on investment, increased the initial investment risk, and deteriorated the cash flow expenditure situation of enterprises, thus affecting the location decisions of MNCs.

MNCs with competitive advantages through capital, technology, brand, and management invest abroad to make full use of overseas resources and to obtain greater profits [3]. Classical location theory considers cost minimization and profit maximization to be important factors affecting the location decisions of MNCs. Classical location theory thus focuses on both the effects of natural resources, labor, technology, and other factor costs as well as the major effect of market size and potential on location decisions [4,5,6]. In addition to production and transportation costs, transaction and information costs are also key considerations for MNCs [7,8]. The cultural and institutional turn in economic geography in the 1990s embedded social, cultural, and institutional concepts in location studies, which largely expanded the scope of location theory and had a profound impact on FDI location studies, especially for emerging economies [9]. Researchers generally believe that FDI tends to flow into countries or regions with a welcoming attitude toward MNCs, a high degree of economic freedom, strong property rights protection, just laws, and a high degree of government integrity [10,11,12,13] and that corruption in a potential host country impedes the entry of MNCs [14]. Some studies have reported that the establishment of special economic zones is the most common approach by which governments attract FDI [15,16]. With the rise of a new economic geography, some studies have focused on explaining the location decisions of MNCs from the perspective of agglomeration economic theory. According to this theory, foreign-funded enterprises can benefit from being in areas with high concentrations of established foreign-funded enterprises, and the knowledge and information spillover from the established foreign-funded enterprises and established industrial connections can help new investors overcome the disadvantages of being outsiders, producing a self-reinforcing effect [17]. Some studies argue that an agglomeration economy is decisive for securing FDI inflow [18].

Land marketization is defined as the process through which the market mechanism comes to play a fundamental and central role in the allocation of land resources [19]. It is a key element of economic system reform in China and in other developing countries. Land marketization not only clarifies land property relations but also affects land prices, thus exerting a complex effect on FDI. However, minimal attention has been paid to the effects of land marketization level and the resulting changes in land prices on FDI in developing countries. Unlike the land markets of other countries, China’s land market is characterized by a clear “dual-track” system. Local governments have an incentive to sell commercial and residential lands at high prices through a market-based system of bidding and auctioning to obtain a greater amount of land mortgage [20], but local governments often “attract investment with land” to achieve political objectives [21,22]. In addition, Chinese cities are characterized by hierarchical relations that reflect regional differences in the development rights and the investment attraction efforts of local governments [23]. Therefore, this study uses data on primary land market transactions in 256 Chinese cities from 2003 to 2019 to study the effect of land marketization level and land prices on FDI and to reveal the heterogeneity of the impact of land marketization level and land prices on FDI in cities of different levels. This study mainly fills in the literature on the effect of the land marketization level in developing countries on FDI in different level cities through land prices. The main innovations are first, reflecting the impact of land marketization level and its resulting land price on FDI in developing countries through empirical analysis, and second, revealing the heterogeneity of the role of land marketization level and land prices on FDI. The findings will enrich the theoretical research on the location decisions of MNCs and may serve as a reference for local governments in developing policies to attract FDI.

The remainder of this paper is organized as follows. Section 2 reviews the effects of land marketization level and land prices on FDI and proposes two hypotheses. Section 3 outlines the city-level classification method, data sources, and methodology employed in this study. Section 4 describes the changes in land marketization level, land prices, and FDI in China and explores the heterogeneity between cities of different levels. Section 5 tests the two proposed hypotheses. Section 6 explores the effects of land marketization level and land prices on FDI in cities of different levels. Section 7 presents our conclusion and discussion.

2. Literature Review

2.1. Effect of Land Marketization Level on FDI

The more effective a market economic system is, the more efficient the allocation of resources and the smaller the transaction costs in various economic actors within that system are. Therefore, all else being equal, regions with high levels of marketization tend to receive the most FDI [24,25,26]. American MNCs prefer to establish branches in areas with strong intellectual property protection laws, minimal government interference in business operations, minimal government corruption, and effective contract enforcement [27]. Japanese MNCs prefer market-oriented cities like Shanghai to politically oriented cities like Beijing for investment in China [28]. Land marketization is a key component in the reform of market economic systems and a crucial initiative to protect the rights and interests of foreign direct investors. Studies show that since 1986, Vietnam’s land marketization level has increased, creating greater peace of mind for domestic and foreign investors [29]. Some studies believe that the market-oriented development of land in China has increased the income of local governments for financing large-scale infrastructure projects and providing public goods, thus improving the urban business environment [30].

2.2. Effect of Land Marketization Level on FDI through Land Prices

In a market economy, land is both a scarce resource and an important tangible asset, the use value and price of which must be fully reflected in the land market [31,32]. As land is a basic input factor for industrial production activities, changes in land prices can significantly affect the production and operating costs of enterprises. Classical location theory suggests that the minimization of production costs is the main factor affecting the location decisions of MNCs [33]. FDI, as the main object of investment attraction (especially in developing countries), enjoys some preferential policies in host countries, especially preferential land policies, such as land allocation without compensation and exemption from urban land-use taxes [34]. For example, in European countries and Korea, local governments often supply land at a low cost to attract out-of-region investment, especially FDI [35,36]. In the early stages of China’s industrialization, low land costs attracted a flood of FDI in manufacturing [37]. However, since the advent of China’s market-oriented economic reforms, the increase in the average land marketization level in China has promoted considerable increases in land prices, thus increasing the cost of industrial land for foreign-funded manufacturing enterprises and causing foreign-funded manufacturing enterprises to lose their cost advantage [38]. In addition, the increase in land price increases the cash flow pressure of enterprises and significantly reduces the currency investment of foreign enterprises. In order to avoid the high cost caused by land prices, foreign enterprises tend to enter the Chinese market through joint ventures, but the number of domestic joint ventures will not increase indefinitely [34].

3. Hypotheses

In summary, as mentioned above, the effect of land marketization level and the resulting land price change on foreign direct investment provides an explanatory basis for this study. We present two hypotheses in this section.

First, we hypothesize that

H1. 

land marketization level is significantly and positively related to FDI. The higher the level of land marketization in a city, the more conducive this is to the influx of FDI.

Second, we hypothesize that

H2. 

the land marketization level significantly restricts FDI by increasing land prices.

4. City Level Classification and Methodology

4.1. City Level Classification

In China, cities can be classified according to their administrative levels, geographical locations, and economic development. China’s administrative division system for cities includes municipalities directly under the central government, sub-provincial cities, non-sub-provincial capitals, and prefecture-level cities. This administrative hierarchy causes the political statuses of and access to resources and development opportunities in different cities to vary widely. In China, municipalities directly under the central government and sub-provincial cities are often given priority as “testing grounds” for the development and early and pilot implementation of economic reforms and opening-up initiatives [39]. These cities have greater access to resources allocated by the central government. The socioeconomic development of non-sub-provincial capitals benefits not only from their political status in the province but also from various economic and land supply policies. Prefecture-level cities are often neglected by higher-level governments and are, therefore, at a distinct disadvantage in terms of socioeconomic development. In this study, municipalities directly under the central government, sub-provincial cities, and non-sub-provincial capitals are classified as high-administrative-level cities, and prefecture-level cities are classified as low-administrative-level cities.

Chinese cities can also be categorized as coastal or inland cities according to their geographical locations. Coastal cities with favorable locations are important nodes in China’s export-oriented national economy and are strongly supported by economic policies. Inland cities do not have these advantages.

In addition, Chinese cities can be categorized as eastern, central, or western according to their natural environments, geographical locations, and regional economic policies. Eastern cities have more favorable natural environments and geographical locations, convenient transportation, and greater access to technology and generally developed earlier than did cities in other regions. Central cities have less favorable location conditions than do eastern cities. Western cities have less economic development level, scientific and technological resources, transportation, and other advantages than do eastern or central cities.

In this study, high-administrative-level cities, coastal cities, and eastern cities are classified as high-level cities, whereas prefecture-level cities, inland cities, and central and western cities are classified as low-level cities. Of the 256 cities examined in this study, 222 and 34 are high- and low-administrative-level cities, respectively; 203 and 53 are coastal and inland cities, respectively; and 100, 95, and 61 are eastern, central, and western cities, respectively.

4.2. Methodology

4.2.1. Calculation Model of Land Marketization Level

China’s land market can be divided into primary and secondary markets. The secondary land market is usually considered a well-developed market in which both parties to a transaction negotiate freely to reach a deal [19]. Therefore, in this study, we do not consider the secondary land market when measuring the land marketization level. The primary land market refers to the market of land-use rights transfer, specifically the exchange of land property rights that occurs between the state as the subject of land ownership and the subject of land-use rights. This exchange is generally reflected in the process of state land transfer and lease of land-use rights for a limited period with compensation. The main transaction methods involved in the primary land market are bidding, auction, listing, agreement, allocation, and lease. The marketization level reflected in different land transaction methods varies. The four main means of land transfer listed in ascending order of marketization are agreement, bidding, listing, and auction [40,41]. Allocation is the state’s delivery of state-owned land use to land users without compensation, and no marketization is involved. Most land acquired through allocation in China is land for state offices, military use, infrastructure, and public welfare. Land that is developed for national institutions and public utilities is usually not the main target of FDI. In addition, leasing is another form of paid use of state-owned land, which is different from land transfer. However, few land leases are involved in primary land market transactions. Therefore, we use the previous methods [19,41] to measure land marketization level according to records of four land transfer methods, specifically by using the following equation:

$${{\displaystyle LM}}_{it}=\frac{({{\displaystyle x}}_{it}\times {{\displaystyle w}}_1+{{\displaystyle y}}_{it}\times {{\displaystyle w}}_2+{{\displaystyle z}}_{it}\times {{\displaystyle w}}_3+{{\displaystyle p}}_{it}\times {{\displaystyle w}}_4)}{({{\displaystyle x}}_{it}+{{\displaystyle y}}_{it}+{{\displaystyle z}}_{it}+{{\displaystyle p}}_{it})}$$

(1)

where LM_it denotes the land marketization level of city i in year t; x, y, z, and p denote the number of parcels for which the corresponding land transaction method is an agreement, bidding, listing, and auction, respectively; and ${{\displaystyle w}}_1$, ${{\displaystyle w}}_2$, ${{\displaystyle w}}_3$, and ${{\displaystyle w}}_4$ denote weights of 0.3, 0.7, 0.8, and 1, respectively, as determined using the Telfer method.

4.2.2. Gravity Center Model

We use the gravity center model to track the overall evolution of the spatial distribution of FDI in China. This method mainly reveals the location of the distribution center of a certain geographical phenomenon and the moving distance and direction in different years [42], which can generally reflect the regional changes in FDI in China. The formula is as follows:

$$X=\frac{{\displaystyle \sum _{i=1}^n{{\displaystyle x}}_i{{\displaystyle a}}_i}}{{\displaystyle \sum _{i=1}^n{{\displaystyle a}}_i}}$$

(2)

$$Y=\frac{{\displaystyle \sum _{i=1}^n{{\displaystyle y}}_i{{\displaystyle a}}_i}}{{\displaystyle \sum _{i=1}^n{{\displaystyle a}}_i}}$$

(3)

where X and Y are the coordinates of the geographical center of the spatial distribution of FDI, a_i is the FDI of city i, x_i and y_i are the coordinates of the geometric center of the administrative boundaries of city i, and n is the number of cities.

4.2.3. Mediation Model

The mediating model consists of a mediating variable, an independent variable and a dependent variable. The relationship among the three variables is as follows. If the independent variable R has a certain influence on dependent variable T through variable S, then S is called the mediating variable of T and T [43]. Therefore, we construct a mediation model (Figure 1) to test the direct and indirect effects of land marketization level on FDI with land prices as the mediating variable according to relevant studies [44]. We test the mediating effect by performing a Sobel Test, combining Equations (4)–(6), which are expressed as follows:

$${{\displaystyle \ln (FDI}}_{it})=\alpha +\phi {{\displaystyle LM}}_{it}+{{\displaystyle \gamma }}_{j}con+{{\displaystyle \epsilon }}_{it}$$

(4)

$${{\displaystyle \ln (LP}}_{it})=\alpha +\chi {{\displaystyle LM}}_{it}+{{\displaystyle \gamma }}_{j}con+{{\displaystyle \epsilon }}_{it}$$

(5)

$${{\displaystyle \ln (FDI}}_{it})=\alpha +{\phi }^{\prime }{{\displaystyle LM}}_{it}+\theta {{\displaystyle \ln (LP}}_{it})+{{\displaystyle \gamma }}_{j}con+{{\displaystyle \epsilon }}_{it}$$

(6)

where FDI_it denotes the FDI in city i in year t; LM_it and LP_it denote the land marketization level and land prices, respectively, in city i in year t; con denotes the control variables; ε denotes the random error term; α denotes the constant term; φ, χ, φ’ and θ denote the estimated coefficients; and γ_j denotes the estimated coefficient of the jth control variable.

Equation (4) tests whether the effect of land marketization level on FDI is significant. Equation (5) tests whether the effect of land marketization level on the mediating variable is significant. Equation (6) tests whether the effect of mediating variable on FDI is significant after controlling for the effect of land marketization level.

To control for the effects of other socioeconomic factors on FDI, we selected the following factors as initial control variables in this study.

(1)

Labor cost and exchange rate. According to cost theory, countries with lower production costs or higher expected profits easily attract investment from MNCs [45,46,47]. Labor costs usually account for a large proportion of production costs, and the impact of labor costs on the production costs of MNCs is similar to that of land prices. Exchange rates can also affect the sales and revenue of MNCs, which, in turn, affects their profits [14]. In this study, the average annual wage of urban workers and the exchange rate of USD to RMB are used to represent labor costs and exchange rates, respectively.

(2)

Market size. According to market theory, market size is a key factor affecting the profitability and development of enterprises. Regions with larger markets are the most favorable locations for enterprises [4,6]. In this study, we use population size to represent the market size.

(3)

FDI agglomeration. Along with the development of a new economic geography, factors such as agglomeration economy and path dependence have started to be incorporated into the theoretical frameworks of the location decisions of MNCs [48,49,50]. Some empirical studies indicate that an agglomeration economy has a significant positive effect on the regional utilization of FDI in the Chinese economy during periods of rapid transition [51,52]. In addition, the long-term path-dependent effect produced by investment agglomeration results in further agglomeration of cross-border investment activities in pre-selected locations because MNCs can gain insight into potential host countries through information provided by previous investors to minimize risk and market uncertainty [53,54]. In this study, we use cumulative FDI to represent investment agglomeration.

(4)

Economic specialization and diversification. Cities with higher specialization tend to be more production-efficient and, therefore, more attractive to foreign investors. Proponents of economic diversification argue that diversified agglomeration economies can provide MNCs with more benefits in terms of shared services, specialized factor markets, upstream and downstream industrial linkages, knowledge spillovers, and industry-relevant information [25]. In this study, the entropy of the sector with the largest employment share and the inverse of the Hirschman–Herfindahl index are used to represent economic specialization and economic diversification, respectively [Appendix A].

(5)

Infrastructure. Effective infrastructure helps minimize production costs for enterprises and is crucial for attracting FDI [27,55]. Road density is often used to measure infrastructure development. However, given the limited data sources and the effect of mobile communication technology on information transmission, we use the number of cell phone subscribers per capita to represent the quality of infrastructure in this study.

Table 1. Variable definitions.

Variables		Definition
Dependent variable: FDI		Foreign direct investment ($10,000)
Key explanatory variables	Land marketization level (LM)	Ratio of the weighted sum of the number of parcels exchanged through the four means of land transfer to the total number of parcels
	Land prices (LP)	Average land prices of parcels exchanged through the four means of land transfer (¥10,000 yuan/Ha)
Control variables	Labor cost (LC)	Annual average wage of urban workers (RMB yuan)
	Exchange rate (ER)	Exchange rate of USD to RMB
	Market size (MS)	Population size (10,000 persons)
	FDI agglomeration (FA)	Sum of FDI from 2002 to year t ($10,000)
	Economic specialization (ES)	Entropy of the sector with the largest share of urban employment
	Economic diversification (ED)	Reciprocal of the Hirschman–Herfindahl exponent
	Infrastructure (MB)	Number of cell phone subscribers per capita (units per 10,000 people)

presents the basic information for each variable. To avoid pseudo-regression, we perform a Harris–Tzavalis (HT) test suitable for short panel data to evaluate the smoothness of the time series of each variable. The results reveal unit roots in market size and FDI agglomeration, FDI agglomeration alone, and market size alone in the variables affecting FDI in high-administrative level cities; coastal and eastern cities; and inland, central, and western cities, respectively, in the grouped regressions. The variables with unit roots in the grouped regressions exhibit first-order differences, and we perform unit root tests accordingly. The results indicate that the first-order difference series is smooth (

Table 2. Unit root tests.

Variables	256 Cities	High-Administrative	Low-Administrative	Coastal	Inland	Eastern	Central	Western
	Z Value
ln (FDI)	−22.85 ***	−5.45 ***	−22.02 ***	−11.72 ***	−21.08 ***	−12.43 ***	−13.22 ***	−16.33 ***
LM	−25.68 ***	−9.49 ***	−24.27 ***	−16.10 ***	−21.19 ***	−18.99 ***	−13.01 ***	−13.60 ***
ln (LP)	−60.34 ***	−22.59 ***	−57.02 ***	−28.64 ***	−53.13 ***	−38.36 ***	−33.48 ***	−32.78 ***
ln (LC)	−54.96 ***	−4.92 ***	−52.86 ***	−31.40 ***	−30.27 ***	−40.04 ***	−23.33 ***	−21.03 ***
ER	−77.61 ***	−28.28 ***	−72.27 ***	−35.31 ***	−69.11 ***	−48.51 ***	−47.28 ***	−37.89 ***
ln (MS)	−4.98 ***	3.24	−9.8 ***	−13.81 ***	5.54	−15.75 ***	3.71	2.38
d.ln (MS)		−30.73 ***			−70.81 ***		−47.55 ***	−42.11 ***
ES	−34.90 ***	−11.39 ***	−32.77 ***	−9.72 ***	−32.95 ***	−17.26 ***	−8.92 ***	−30.08 ***
ED	−10.84 ***	−3.15 ***	−10.66 ***	−4.49 ***	−9.82 ***	−5.50 ***	−5.61 ***	−9.22 ***
ln (FA)	−5.08 ***	0.75	−5.23 ***	0.60	−6.09 ***	3.29	−2.28 **	−9.56 ***
d.ln (FA)		−15.41 ***		−14.17 ***		−21.97 ***
ln (MB)	−30.80 ***	−15.98 ***	−27.20 ***	−14.12 ***	−29.59 ***	−17.19 ***	−18.25 ***	−22.64 ***

Note: *** p < 0.01, ** p < 0.05.

).

4.3. Data Sources

The data used to measure land marketization level and land prices in this study are mainly derived from published statistics and land supply transaction announcements. Among them, the prices and parcel amounts obtained through bidding, auction, listing, and agreement transactions from 2003 to 2008 are mainly from the China Statistical Yearbook of Land and Resources. However, after 2008, the National Bureau of Statistics stopped publishing the breakdown of bidding, auctioning, listing, and agreement data in the Yearbook. Fortunately, the “Specification for the Sale of State-Owned Land Use Rights by Bidding, Auction, and Listing (for Trial Implementation)”, which came into effect on 1 November 2007, stipulates that local land authorities must announce the results of each land transaction on China Land Market Online (Land Supply Result Notice: available online: https://www.landchina.com/ (accessed on 18 December 2021)). Each announcement must specify the electronic supervision number, administrative district, location, area, and use patterns of the land, as well as the supply means, transaction price, and contract signing date, among other details, for each transaction. We obtained 1,987,674 transaction records from 2009 to 2019 by using web crawler technology. After the records for land transactions in non-sample cities, for allocated and leased land, and with incomplete information were excluded, 1,155,679 transaction records remained. Among these records, 5068, 111,475, 614,094, and 425,042 are for bidding, auction, listing, and agreement, respectively. Additional socioeconomic data used in this study are from the China City Statistical Yearbook.

5. Temporal Changes and Heterogeneity in Land Marketization Level, Land Prices, and FDI

5.1. Land Marketization Level

The results of the land marketization level calculation model show that from 2003 to 2019, the average land marketization level in 256 cities followed an overall increasing trend, with a total increase of 47.26% (Figure 2), which illustrates how land marketization in China continues to increase as China’s market economic system reforms continue to progress. Although the Regulation on the Right to Use State-Owned Construction Land by Bidding, Auction, and Listing was passed at the Fourth Ministerial Meeting of the Ministry of Land and Resources on 3 April 2002, the regulation has only exerted a weak positive effect on land marketization. The revision of this regulation was approved at the third Ministerial Meeting of the Ministry of Land and Resources on 21 September 2007 and has played a major role in promoting land marketization in China. The average land marketization level increased significantly from 2006 to 2008.

Based on the results of the land marketization level calculation model, we compare statistics on the average value of the land marketization level of the high and low administrative levels, coastal and inland cities, eastern, central and western cities. The results show the trends in land marketization levels in different classes of cities are similar to the overarching trend in land marketization among all 256 cities but exhibit significant regional heterogeneity. The average land marketization level in high-administrative-level cities was significantly lower than that in low-administrative-level cities from 2003 to 2008, but the difference was nonsignificant from 2009 to 2019. The average land marketization level in coastal cities was significantly lower than that in inland cities from 2003 to 2007, but the difference was nonsignificant from 2008 to 2019. Central cities had a higher average land marketization level than did cities in other regions from 2003 to 2007, but the difference among the cities in the three major economic regions narrowed from 2008 to 2019. During the 17-year study period, the average land marketization level of high-administrative-level cities was 0.61, which was lower than that of low-administrative-level cities. The average land marketization level of coastal cities was 0.61, which was lower than that of inland cities. The average land marketization level of eastern cities was 0.63, which was lower than that of central cities but higher than that of western cities. In general, land marketization levels in high-level cities have increased significantly but generally remain lower than those in low-level cities.

5.2. Land Prices

With the increase in land marketization in China, the land pricing mechanism is playing an increasingly important role in the allocation of land resources, and land transfer prices have increased significantly as well (Figure 3). The average price of land transfer among the 256 cities fluctuated from 2003 to 2019, and the average price of land transfer in 2019 was 5.05 times higher than that in 2003. From the perspective of temporal variation in land transfer prices in cities of different levels, the average land transfer prices of high-administrative-level cities are always higher than those of low-administrative-level cities, coastal cities are always higher than those of inland cities, and eastern cities are always higher than those of central and western cities. The increases in the average price of land transfer differed significantly between high- and low-level cities. The average land transfer prices of high-administrative-level, low-administrative-level, coastal, inland, eastern, central, and western cities in 2019 were 6.96, 4.38, 6.67, 4.41, 5.96, 4.47, and 3.77 times the corresponding average land prices in 2003, respectively. Therefore, the land transfer prices of high-level cities are higher and are increasing more rapidly than those of low-level cities.

5.3. FDI

The total amount of FDI in the 256 cities increased rapidly from 2003 to 2014 and then declined slightly from 2014 to 2019 (Figure 4). Overall, the total amount of FDI increased faster, with an average annual growth rate of 7.71%. The temporal variation in the proportion of FDI in high- and low-level cities was significant. From 2003 to 2019, the proportions of FDI in high-administrative-level, coastal, and eastern cities all decreased to varying extents. The proportion of FDI in high-administrative-level cities alternated with that in low-administrative-level cities from 2003 to 2007 but remained lower than that in low-administrative-level cities from 2008 to 2019. The proportion of FDI in coastal cities was higher than that in inland cities from 2003 to 2006, but the opposite was true from 2007 to 2019. Although the gap between the proportions of FDI in eastern cities and in central and western cities has narrowed significantly, the proportion of FDI in eastern cities remains higher than those in central and western cities. This indicates that low-level cities have become the “new hotspots” for FDI. The change in the spatial distribution of FDI further supports this trend (Figure 5). The gravity center of the spatial distribution of FDI in China moved gradually toward the northwest from 2003 to 2017 and then rapidly toward the southwest from 2017 to 2019. The geographical center of the spatial distribution was in Wuhu City, Anhui Province, in 2003 and in Xinyang City, Henan Province, in 2019, moving approximately 280 km to the northwest. Although the proportion of FDI in high-level cities has decreased, the average amount of FDI in such cities is still higher than that in low-level cities. The average amount of FDI in 2019 in high-administrative-level cities was 3.66 times that in low-administrative-level cities, the average FDI in coastal cities was 2.30 times that in inland cities, and the average FDI in eastern cities was 2.11 and 2.22 times those in central and western cities.

6. Effects of Land Marketization Level and Land Prices on FDI

To ensure the reliability of the mediation analysis results, we improve the robustness of the regression analysis by changing the control variables in the equation. The results reveal that market size and economic specialization can be used as reasonable control variables. As indicated in Figure 6, land marketization significantly promotes FDI, which is in line with Hypothesis 1. Land marketization, which is oriented by market demand and helps achieve the rational allocation of land resources, is a key part of China’s economic market-oriented reforms. Land marketization in China, which aims to build an open, fair, and equitable land resource allocation mechanism, protects the land needs and property rights interests of MNCs. Land marketization also encourages local governments to take the initiative to support economic development and to proactively adopt various preferential policies to attract FDI, thereby promoting the transformation and advancement of industry. Therefore, in the context of capital shortages and technological backwardness, FDI is regarded as an important tool for local governments in China to fill the capital gap and promote technological advancement in the pursuit of rapid economic growth.

Land price changes caused by the evolution of land marketization exert a significant positive effect on FDI in China, contrary to Hypothesis 2. The intensity of the indirect effect of land marketization level on FDI through land prices exceeds the direct impact of land marketization level on FDI. Because land is a scarce resource, increases in urban land prices can reflect land scarcity, which enhances the role of land property as an investment vehicle. Considering the possibility of future land appreciation in a given city, MNCs may elect to directly invest in the city to obtain high land premiums [56]. In addition, both domestic and foreign-funded enterprises face considerable financing constraints. Using land as “collateral”, in anticipation of rising land prices, enterprises can hoard land, mortgage it to banks when necessary and break their financing constraints through mortgage loans, which is crucial to their long-term development [38,57]. Therefore, the collateral nature of land also determines increases in urban land prices to a certain extent, thus affecting FDI inflows. However, rising land prices do not always help attract FDI. As illustrated in Figure 7, land prices and FDI are not simply negatively correlated but rather have a nonlinear inverted U-shaped relationship.

7. Heterogeneity in the Effects of Land Marketization Level and Land Prices on FDI

As discussed in Section 4.1 and Section 4.2, land marketization levels and land prices differ significantly between low-and high-level cities in China. Therefore, we use the mediating model to estimate the regional heterogeneity of the influence of land marketization level and land price on FDI in high- and low-administrative-level, coastal and inland, eastern, central and western cities, respectively. The results of the group regression further support the robustness of the effects of land marketization level and land price on FDI.

The regression results reveal that land marketization promotes FDI in both high- and low-administrative-level cities but that the positive effect of land marketization on FDI in low-administrative-level cities is stronger than that in high-administrative-level cities (Figure 8). A possible reason for this trend is that low-administrative-level cities have abundant land resources and high land marketization levels, which is conducive to attracting MNCs. The mediating effect of land prices on FDI in high-administrative-level cities is significantly stronger than that in low-administrative-level cities, and the ratio of the mediating effect to the total effect in high-administrative-level cities is also higher than that in low-administrative-level cities. A possible reason for this trend is that because high-administrative-level cities have higher levels and a more rapid pace of economic development, greater land scarcity, and more rapidly increasing land prices than do low-administrative-level cities, the role of land properties as investment goods is more prominent in high-administrative-level cities. In addition, high-administrative-level cities have higher land prices, which enables them to obtain higher land revenue from land transfer. This additional land revenue helps local governments invest in infrastructure, including transportation, post and telecommunications, and water and electricity infrastructure, as well as urban green spaces. Improvements in infrastructure improve the business environment in a city and make the city more attractive to potential investors and residents, which, in turn, attracts FDI.

Land marketization level is significantly related to FDI in both coastal and inland cities (Figure 9). However, the direct positive effect of land marketization level on FDI in coastal cities is significantly lower than that in inland cities. The positive impact of land marketization level on FDI in low-administrative-level cities is stronger than in high-administrative-level cities, which may explain this estimation. Although the mediating effect of land prices in the relationship between land marketization level and FDI in coastal cities is lower than that in inland cities, the mediating effect accounts for a greater proportion of the total effect in coastal cities than it does in inland cities. A possible reason for this trend is that in addition to the prominent role of land properties as investment vehicles and the favorable business environment of coastal cities, coastal cities are key nodes in China’s export-oriented economy and are the most strongly supported by economic policies. These cities often have a more complete industrial chain as well as numerous skilled workers and serve as the centers of export-oriented processing and manufacturing industries.

The estimated coefficient of the direct effect of land marketization level on FDI in eastern cities is negative and nonsignificant (Figure 10). This may be because, during the period of rapid industrialization, population and economic activities rapidly gathered in eastern cities, resulting in an uneconomic agglomeration effect and an overall weakening of the initial location advantage. Land marketization level is significantly and positively related to FDI in central and western cities, especially central cities. This may be because eastern cities are victims of a spillover effect due to uneconomical agglomeration whereby FDI is moved elsewhere, with central and western cities (especially central cities adjacent to the eastern region) benefiting from this spillover of foreign enterprises. Land marketization exerts a significant positive effect on FDI in eastern, central, and western cities through land prices. The mediating effect of land prices is significantly stronger in eastern cities than in central and western cities, and the role of land properties as investment vehicles is more prominent. The overall increase in land prices in central cities is not as drastic as that in western cities, and the role of land properties as investment goods, as well as the mediating effect of land prices, are also less prominent.

8. Conclusions and Discussion

Using data on primary land market transactions in 256 Chinese cities from 2003 to 2019, we analyzed the effects of land marketization level and land prices on FDI in the context of land system reform. The results of our study reveal heterogeneity in both the direct and indirect effects of land marketization levels on FDI.

(1)

With the continuous improvement of China’s market economic system, the level of land marketization in China has also increased, with significant regional heterogeneity. Although the land marketization level in high-level cities is not higher than those in low-level cities, they are increasing at a more rapid pace. With the increase in land marketization, land prices have also increased significantly, especially in high-level cities. Although the relative attractiveness of high-level cities to foreign investors has declined over time, high-level cities are still major destinations for FDI. The improvement of the land marketization level and the rise of land prices have promoted the inflow of FDI into China. The positive effect of land marketization level on FDI in high-level cities is weak, and land marketization even negatively affects FDI in some cases. By contrast, the land marketization level generally exerts a stronger positive effect on FDI through land prices in high-level cities than it does in low-level cities.

(2)

Some studies have argued that, in the context of imperfect land marketization, some local governments in China still use their administrative power to control the land market to attract FDI and expand their tax bases [58], especially in central and western economic development zones [21]. However, the results of the present study reveal that land marketization can significantly promote FDI in China because market-oriented land development is a key component of market-oriented economic reforms in China and an important safeguard for the rights and interests of foreign direct investors, which further confirms the positive impact of property rights protection and contract respect on the market economy [59,60]. Traditional location theory suggests that land cost is an important factor influencing the location of a firm [61]. Multinational corporations chose to invest in China mainly because of the cheap land and labor and the large market size [62]. Since entering the 21st century, China’s production costs have been rising faster, which has a certain negative impact on FDI. For example, the high land cost is a negative factor in attracting FDI in the Pearl River Delta and Yangtze River Delta cities [63]. Overall, however, the increases in land prices caused by China’s special land market system do not significantly inhibit FDI but rather significantly promote FDI, especially in high administrative-level, eastern and central cities. It is worth noting that Increasing land prices do not always help attract FDI; after land prices reach a certain threshold, continued increases in land prices may discourage FDI. These findings partially reflect the effects of land property rights protection and land prices on FDI in developing countries within the context of economic system reform, thereby enriching the literature on the theoretical bases of the location decisions of MNCs and elucidating the mechanisms underlying the heterogeneity in the spatial distribution of FDI within China, which can provide a reference for the macro-level location choice of FDI in China and for the central government to coordinate regional development and adjust FDI utilization policies. However, there are still shortcomings in this study. Namely, the sectors of FDI have not been subdivided, and the impact of land marketization and land prices on FDI in traditional manufacturing industries, sophisticated manufacturing industries, traditional service industries and modern service industries has not been explored.

(3)

The differences in land marketization levels and land prices between high- and low-level cities in China strongly affect the regional distribution of FDI. Although the amount of FDI in low-level cities has increased more rapidly than that in the high-level study, high-level cities are still the primary destinations of FDI, which strongly affects the balance in China’s economic development. China’s current economic development efforts are focused primarily on high-quality development and are expected to remain so for the foreseeable future. Coordinated development is essential for ensuring high-quality economic development. Therefore, the central government must continue prioritizing efficiency and equity, strengthening land market reform, gradually improving the FDI management system, and promoting the coordinated opening up of the central, western and coastal regions. High-level cities should explore land management innovation, strengthen the development of inefficient construction land, and provide land security for FDI in sophisticated manufacturing, strategic emerging industries, digital economy and other fields. Low-level cities should seize the opportunity of industrial transfer, give play to the comparative advantages of land and labor costs, deepen the reform of the market system and investment environment, and create a fair competition environment for foreign-funded enterprises.