How the Marketization of Land Transfer Affects High-Quality Economic Development: Empirical Evidence from 284 Prefecture-Level Cities in China

Abstract

The allocation of urban land from planned to market-oriented is an important part of China’s economic market-oriented reform, but its impact on high-quality economic development still lacks direct testing. Based on the data of prefecture-level city panels from 1999 to 2019, this paper analyzes the impact mechanism and effect of land transfer marketization on the high-quality development of urban economy by constructing multiple land transfer marketization indicators. The study found that the marketization of land transfer has a significant role in promoting high-quality economic development in the long run. The specific mechanism is that the marketization of land transfer affects the high-quality development of the economy through the financing effect and the resource allocation effect. The degree of marketization of land transfer can be increased, which can not only promote the expansion of production scale by increasing the degree of land capitalization and increasing the scale of urban financing, but also improve the efficiency of resource allocation by giving more effective play to the land price signal and guiding the combination of production factors to match more effectively. However, this paper also finds that the effect of land financing has a very complex impact on resource allocation, and the impact of financing in the primary and secondary land markets on the efficiency of resource allocation is generally completely different. The research results of this paper have rich policy implications and have practical reference value for evaluating and improving the current urban land transfer system. In the future, we should continue to improve the land transfer system in the direction of marketization, reduce the improper administrative interference of local governments in land transfer, improve the level of marketization of the primary land market, and further develop the secondary land market.

1. Introduction

Over the past 40 years since the implementation of reform and opening up, China has made major achievements in economic construction, laying a solid foundation for high-quality economic development. China’s GDP increased from US$0.1495 trillion in 1978 to US$101.5986 trillion in 2020, an increase of about 275 times, and its total economic volume jumped from fifteenth place in the world to second place in the world. China’s per capita GDP increased from $156 in 1978 to $10,500 in 2020, entering the ranks of the middle-developed countries. However, in the process of China’s rapid economic growth, the traditional demographic dividend has faded, the overall return on capital has declined, and the ability to innovate independently is weak. Meanwhile, there has been an accumulation of financial risks, an intensification of resource and environmental constraints, a widening gap in residents’ income distribution, and a middle-income trap, all of which have attracted great attention from scholars and policy makers. At the same time, after experiencing a long period of rapid growth, the main contradictions of Chinese society and the phased characteristics of economic development have undergone fundamental changes. It is precisely based on the profound changes in China’s economic development environment, and the fundamental change in the traditional development model accumulated by “paving the way” and “projects” in the past, that the 19th National Congress of the Communist Party of China has outlined the major thesis that China’s economy has shifted from a stage of high-speed growth to a stage of high-quality development. The 2018 Central Economic Work Conference once again stressed the need to “adhere to the general tone of the work of seeking progress in stability, adhere to the new development concept, and adhere to promoting high-quality development”. In 2020, on the basis of the domestic development situation and in grasping the general trend of international development, the central government further proposed to “accelerate the construction of a new development pattern with the domestic cycle as the main body and the domestic and international dual cycles promoting each other”, which put forward a new strategic concept for the high-quality transformation of China’s economy. High-quality development has become an inevitable requirement for adapting to the changes in the main contradictions of society and achieving the sustained and stable development of China’s economy.

The efficiency of resource space allocation is an important factor affecting the high-quality development of the economy. In terms of factor resource allocation efficiency, scholars clarify the meaning of resource allocation, i.e., how to improve the efficiency of resource utilization under given limited resources in neoclassical economics, by stating that, under all resource allocation methods, the market is effective and that the perfectly competitive market can most effectively guide the allocation of social resources; though this does not exclude the important role of the rule of law and private property rights protection [1]. Furthermore, scholars have also conducted in-depth research on the quality effect of economic growth in the efficiency of capital and labor allocation [2,3,4]. However, research on the impact of land allocation is relatively scarce. In the new stage of China’s economic development, the allocation efficiency of land resources has an important impact on high-quality economic development [5,6]. Since its reform and opening up, China’s urban land supply system has undergone many major reforms, which provides unique conditions for studying the high-quality economic development effect of land element allocation. During the planned economy period, all urban land in China was allocated by the government through gratuitous allocation. Since the mid-1980s, China’s urban operational construction land has gradually shifted from gratuitous allocation to paid transfer. In 2001, the reform of China’s urban land transfer system was fully opened, in 2004, the transfer of commercial and residential land was fully auctioned and listed, and the transfer of industrial land in 2007 also required bidding, auctioning and listing. With this, the transfer of China’s urban operational construction land officially entered the stage of comprehensive marketization. The marketization of urban land transfer is an important part of China’s economic market-oriented reform, which has had a far-reaching impact on China’s economic development since the 21st century and has aroused the exploration of many scholars [7,8]. What kind of impact does the marketization of land transfer have on the high-quality development of China’s economy? The discussion in this regard is mainly a theoretical discussion, and there is a lack of empirical research. Some scholars believe that the marketization of land transfer has brought about a rich “land system dividend”, which is conducive to improving the efficiency of land resource allocation, thereby promoting high-quality economic development [9,10,11]. However, some other scholars believe that the market-oriented transfer of land obtained by the high-prices has brought about problems of land finance and high housing prices, which have had a negative impact on the quality of social and economic development and people’s livelihoods [12,13,14]. It can be seen that there is still a lack of unified understanding of this issue in the academic community. At the same time, there is also a lack of systematic sorting out and empirical evidence on the impact mechanism of land transfer marketization on high-quality economic development, which may be an important reason for the divergences in the impacts on the marketization of land transfer.

In order to make up for the lack of relevant research, this paper first discusses the mechanism of the marketization of land transfer affecting the high-quality development of the economy, and then empirically estimates the impact of the degree of land transfer marketization on the high-quality development of the urban economy based on panel data of cities at the prefecture level and above in China from 1999 to 2019. It also examines the influence mechanism. This paper uses a dynamic panel econometric model based on GMM to overcome the indigenousness between variables. In terms of measuring key variables, such as land transfer marketization indicators and high-quality economic development, this paper also tries a variety of measurement methods to increase the robustness of the study. Based on the results of different model settings and a series of robustness tests, this paper unanimously concludes that the increase in the degree of marketization of land transfer has a significant role in promoting high-quality economic development in the long run. It is further found that the marketization of land transfer can promote high-quality economic development through financing effects and resource allocation effects. However, it is also found that land financing has a very complex impact on resource allocation efficiency, and that the overall impact of land primary and secondary market financing on allocation efficiency is completely different.

The marginal contribution of this paper reflects the following three aspects: (1) The impact effect of the degree of marketization of land transfer on high-quality economic development was tested by empirically testing the data of China’s prefecture-level cities, which makes up for the lack of empirical evidence and the shortcomings of the research scale of relevant research and broadens the research on the quality of economic growth of land transfer marketization; (2) from the perspectives of financing effects and resource allocation effects, the impact mechanism of land transfer marketization on urban economic growth was tested, and the interaction between the two is discussed; and (3) China’s special land transfer system reform as a quasi-experimental environment has promoted literature research on the economic effects of the land system, and also provided reference for the high-quality economic development of other developing countries around the world.

The remainder of this article is arranged as follows: the second part is a theoretical analysis and research hypothesis; the third part describes the study design, which contains the econometric model and data description; the fourth part contains the measurement results and discussion; and the fifth part contains the conclusions and policy recommendations, as well as the research gaps and prospects.

2. Theoretical Analysis and Research Assumptions

2.1. Market-Oriented Reform of Land Transfer in China

After 1949, China’s land system gradually formed a dual system in which urban land was owned by the state and rural land was collectively owned. During the planned economy, all urban land was allocated and supplied by the government free of charge. After the reform and opening up, on the basis of pilot projects in Shenzhen, Shanghai and other cities, the Land Administration Law promulgated by the central government in 1986 legalized the paid transfer of land use rights, and China’s urban operational construction land gradually changed from gratuitous allocation to paid transfer. In the 1990s, China’s urban land was paid for, mainly by agreement. Due to the lack of openness and transparency in the land supply of the agreement, there is a lot of corruption and rent-seeking space [15,16,17,18,19,20,21,22], and it leads to wasted land use and inefficient use. Compared with commercial and residential land, the order of industrial land transfer is more chaotic [23,24,25,26]. In China’s regional economic competition, local governments have attracted investment with low or even free industrial land transfers and negative land prices-subsidies for infrastructure facilities, which some scholars have called “attracting investment with land” or “seeking development with land” [27,28,29]. As a result, the price of industrial land transfer is very low nationwide, and the waste of industrial land use is very serious [30,31]. In order to curb corruption in the field of land transfer, reduce the loss of related state-owned assets and protect cultivated land, promote the intensive use of land, and also to cooperate with the reform of the economic system, the central government has promoted further market-oriented reform of land transfer.

Marked by the “Vigorous Promotion of Bidding and Auction and Transfer of State-owned Land Use Rights” proposed by the State Council in April 2001 “Notice on Strengthening the Management of State-owned Land Assets” (Guo Fa (2001) No. 15), the market-oriented reform of land transfer in China in the 21st century was launched. In the first phase, the landmark document was the Provisions of the Ministry of Land and Resources on the Transfer of State-owned Land Use Rights by Tender, Auction and Listing (Order No. 11) in May 2002. The document clearly stipulates that “all kinds of commercial land such as commerce, tourism, entertainment and commercial housing” must be transferred in the form of bidding, auction and listing, which will be implemented from July 2002. In March 2004, the Ministry of Land and Resources and the Ministry of Supervision jointly issued the Notice on Continuing to Carry Out the Law Enforcement supervision of the Bidding, Auction, Listing and Transfer of Commercial Land Use Rights (i.e., Order No. 71), which requires that, after 31 August 2004, the commercial construction land for commercial, tourism, entertainment and commercial housing can only be transferred through bidding, auction and listing, the so-called “8.31” limit. In the second stage, the focus of the central government’s land transfer market-oriented reform shifted to industrial land. In August 2006, the State Council issued the Notice of the State Council on Issues Related to Strengthening Land Regulation and Control (Guo Fa (2006) No. 31 Document), proposing to “establish a unified system for the publication of the minimum price standard for the transfer of industrial land” and “industrial land must be transferred by way of bidding, auction and listing”. In September 2007, the Ministry of Land and Resources revised and re-promulgated the Provisions on the Transfer of the Right to Use State-owned Construction Land by Bidding, Auction and Listing (i.e., “Order No. 39”), which clearly stipulates that industrial land (excluding mines) and other lands used by two or more intended land users of the same parcel shall also be transferred by way of bidding, auctioning and listing, which will be formally implemented from November of the same year. From the perspective of the market-oriented reform process of China’s urban land supply, it is mainly a gradual process promoted by the central government from top to bottom. However, the degree of implementation of the market-oriented requirements for land transfer varies greatly from place to place, which provides a rich basis for differentiation analysis for the research in this paper.

After the reform of the tax-sharing system in 1994 and the income tax-sharing reform in 2002, the central government’s fiscal power was relatively concentrated, and the ability to transfer payments was greatly improved. In this context, local governments are facing the dilemma of shrinking financial power and expanding power, and the financial pressure has increased, forcing local governments to seek extra-budgetary revenue to ensure local government economic development and social public service investment needs. Coupled with the central government’s relaxed supervision of land transfer fees, land transfer revenue has become the largest source of extra-budgetary income for local governments. With the continuous rise of financing platforms built by local governments with land use rights as an opportunity, land financing now plays an important role in supplementing local government finances and supporting urbanization development. Since the implementation of the land market-oriented reform system (“Bidding, auctioning and listing” refers to the abbreviation of the methods of bidding, auction and listing and transfer of state-owned land use rights, and the abbreviations are used in the following paragraphs.), land transfer revenue has become an important source of revenue for local government finance in China. Fiscal statistics show that the scale of China’s land transfer revenue shows a rapid expansionary trend, and in 2020, China’s land transfer revenue reached 8.414 trillion yuan, an increase of 15.90% year-on-year, accounting for 84.03% of local fiscal revenue, setting a 33-year record. Undoubtedly, such a huge amount of land transfer income has increased the financial resources of local governments and alleviated the financial pressure brought about by the reform of the tax-sharing system, but the expansion of land revenue scale has also caused a serious mismatch of resource space, which has a deep negative effect on industrial structure upgrading and innovative development, and has become a problem that cannot be ignored in the process of comprehensively promoting high-quality economic development.

2.2. Mechanism Analysis and Research Hypothesis

Through the above analysis, it can be seen that the impact mechanism of land transfer marketization on economic growth is very complicated. The marketization of land transfer will increase the amount of land transfer fees and land mortgage financing [32,33,34,35], bring about the expansion of urban financing scale, and thereby promote the growth of economic output scale, something which is called “financing effect” in this paper. The marketization of land transfer will also have an impact on the allocation efficiency of land resources, something which is called the “allocation effect”.

The “financing effect” is a very important mechanism for the marketization of land transfer to affect the high-quality development of the economy. The marketization of land transfer will increase land and housing prices and improve the financing capacity of the government and enterprises. The rise in land prices has increased the income from land transfers of local governments, and, as the owners of urban land, the ability of land mortgage financing has been enhanced. For enterprises, when there is financial friction, land prices and housing prices rise to increase the mortgage value of their own land and real estate, which can alleviate the financing constraints of enterprises [36,37,38,39,40]. Caballero and Krishnamurthy (2006) argue that, in markets where financial investment instruments are underdeveloped, asset bubbles generated by rising house prices are beneficial to economic development, reducing capital outflows, expanding credit, and increasing investment by providing store-of-value targets [41]. Therefore, the marketization of land transfer can ease the financing constraints of cities, enhance financing capabilities, expand the scale of credit, and promote increasing investment. According to the growth equation, increased investment leads to an increase in capital formation, which in turn leads to an increase in output. A large number of documents on land finance also demonstrate that local governments’ access to land transfer fees and land financing not only makes up for the financial gap in infrastructure construction, but also subsidizes industrialization and realizes “land for development” based on “land financing” [42,43,44,45,46,47,48,49,50]. Based on the above analysis, this article proposes:

Hypothesis 1.

The increase in the degree of marketization of land transfer will help ease financing constraints and expand the scale of urban financing, thereby promoting high-quality economic development.

The allocation effect of land transfer marketization means that the marketization of land transfer can improve the allocation efficiency of land elements, and then improve the allocation efficiency of the economy as a whole. Factor allocation theory states that market-oriented allocation has two major advantages: The competitive pricing of market-oriented transfers has a screening effect and can ensure that efficient enterprises are stationed In addition, the market is more responsive than artificial planning, and can reach an equilibrium between supply and demand more quickly, reducing resource idleness or insufficient supply. Second, market-oriented public bidding can reduce rent consumption caused by rent-seeking. Through market allocation, information is open and transparent, reducing rent-seeking space, reducing the “rent dissipation” of the rent-seeking process or the meaningless loss of other social resources.

The market-oriented reform of land transfer can also have two positive effects. On the one hand, it can reduce rent-seeking corruption activities and reduce the loss of economic efficiency. Tao Kunyu et al. (2010) used provincial panel data from 2003 to 2007 to empirically conclude that the marketization of land transfer can reduce rent-seeking space and reduce the occurrence of illegal cases such as corruption [51]. Cai et al. (2013) also found in the study that, as the degree of marketization of land transfer methods increases, the level of corruption in government land transfer decreases. Furthermore, more potential buyers promote the optimal matching of land resources with land users through open market competition, thereby improving the efficiency of resource allocation [52]. Wang Keqiang et al. (2013) compared the differences in the output flexibility of land elements among enterprises in a development park that obtained land either through marketization or other means and found that the output efficiency of the former type of enterprises was higher [53]. Li Lixing et al. (2016) found that the misallocation of land resources has a significant impact on the productivity of industrial enterprises [54]. Based on these discussions, this article proposes:

Hypothesis 2.

The increase in the degree of marketization of land transfer will help improve the efficiency of urban production resource allocation, thereby promoting high-quality economic development.

However, by influencing the capital price signal to change the allocation combination of urban resource factor inputs, the financing effect generated by the marketization of land transfer will also affect the allocation effect. On the one hand, the financing effect has the potential to strengthen the allocation effect. If land financing is mainly used to expand urban infrastructure and increase productive investment and research and development of enterprises, the financing effect not only plays a role in alleviating the constraints of production financing, but also promotes the improvement of resource allocation efficiency [55], then the impact of the financing effect on the resource allocation effect is positive. Good urban infrastructure and more government subsidies can attract better quality enterprises, promote industrial agglomeration of externalities, and improve the total factor productivity of the economy. Existing enterprises in cities have also promoted productivity by easing financing constraints and increasing productive investment and research and development. However, the financing effect may also weaken the allocation effect. Yu Yongze and Zhang Shaohui (2017) found that the rise in urban housing prices will lead to distortions in the investment structure, and the “squeeze effect” of innovation funds inhibits the level of technological innovation in the region [56]. Bleck and Liu (2018) also revealed that if too much liquidity is injected into the economy, not only will sectors with lower financial frictions such as the real estate sector overheat, but also squeeze out liquidity in sectors with higher financial frictions such as manufacturing. Further, because of the feedback between liquidity inflows, asset prices, and the value of real estate collateral, the crowding out effect also presents as a “self-reinforcing spiral effect” [57]. At the same time, the vicious circle between excessive investment in real estate and credit expansion is likely to trigger credit bubbles, real estate financialization bubbles and other hazards, which is not sustainable for China’s economic growth, so the Chinese government will limit the purchase policy and the pilot reform of real estate tax [58]. According to these documents, if the credit funds generated by land financing are used inappropriately, then the more adequate the supply of credit funds and the lower the cost, the more production enterprises may flock to real estate, squeezing out investment and enterprise research and development in the real economy and thereby reducing the efficiency of resource allocation and having a negative impact on total factor productivity. Based on these discussions, this article proposes:

Hypothesis 3.

The land financing effect produced by the marketization process of land transfer may have both positive and negative effects on resource allocation, and the specific impact is related to how the funds for land financing are applied.

3. Research Design

3.1. Data Sources

The research sample of this paper includes prefecture-level cities and municipalities directly under the central government in China, and the urban scope was the whole city (including municipal districts and non-municipal districts), hereinafter referred to as cities. Due to the lack of relevant economic data, the sample did not include four cities: Chaohu City, Bijie City, Tongren City and Lhasa City. The land transfer data come from the statistical data section of the China Land and Resources Yearbook over the years. Since the city’s complete secondary market transfer and mortgage data were published until 2008, the analysis period for land financing, including the secondary market, was from 1999 to 2008. In 2013, the land mortgage situation of 84 key cities was announced, and the land mortgage data indicators were also included. The economic data at the city level were derived from the “China Urban Statistical Yearbook”, EPS global database, and the wind information database. The GDP and investment data were converted to the constant price of 1996, in order to ensure the stability of the data and reduce the interference of heteroscedasticity, the variable data are logarithmic processing.

3.2. Variable Design

(1) Main variables. A measure of the degree of marketization of land transfer. How to scientifically and reasonably measure the degree of marketization of land transfer is a core basis for the research work of this paper. Many previous documents have undertaken exploratory work on the measurement of the degree of marketization of land transfer [33,59,60]. Summarizing the relevant literature, there are currently two main methods: proportional method and weight method. (1) Proportional method: the proportion of land transferred in the form of bidding, auctioning and listing to the total land transferred is calculated. Considering that the land supply not only has the transfer (including the agreement transfer and the bidding, auction and listing transfer), but also other methods such as allocation and leasing, the proportional method also has two corresponding transfer proportional methods and a supply proportional method. However, the land supply of non-transfer methods is generally non-operational construction land such as public land, educational land, and infrastructure land, and since this paper mainly examines the impact on economic growth, it is believed that the use of the transfer ratio method can better explain the research problems. In addition, the indicator can be used to account for both the number of cases and the proportion of area. Previously, most of the literature used the number of cases as an indicator, so the benchmark index of land transfer marketization in this paper is calculated by the number of cases. (2) Weight method: the measurement of market-oriented indicators relative to the proportional method only considers the deviation between the land supply method and the market-oriented standard, and the weight method also considers the deviation between the actual land supply price and the market-oriented standard. The assumption is that if local governments intervene in the transfer of land, the extent of this intervention can be reflected in the degree of deviation from the price of the land transfer from its potential market value. Much literature has used the weight method to measure the degree of marketization of land transfers [33,59,60]. The specific idea of the weight method is to use the land supply method and the price weighted average to calculate the land marketization index. The calculation formula is $R_t={\displaystyle \sum _{i=1}^n{Z}_i{f}_i}/{\displaystyle \sum _{i=1}^n{Z}_i}$. Among these, $Z_i$ is the land supplied by various transaction methods in the urban land market, and the price weight of various transaction methods. The calculation of price weights is calculated based on the ratio of the actual transaction price of a certain market transaction price to the benchmark transaction price of other trading methods. Liu et al. (2016) used the benchmark of the auctioned land price [33]. However, due to the fact that, after 2008, the “China Land and Resources Yearbook” no longer publishes the sub-data of bidding, auction and listing and transfer, in practice, various localities prefer the listing method of transfer, and the proportion of auction and bidding is very small, we did not calculate the price weight of bidding, auction and listing transfer separately, but took the average price of land for bidding, auction and listing in each city as the benchmark. In terms of the time dimension of price weights, Qian and Mou (2012a) used fixed price weights to calculate the national average price of different land supply methods from 2003 to 2008 [60]. However, we calculated the price weight of the change in cities by using the average prices of different supply methods in each city each year.

Based on the above discussion, we constructed eight indicators to measure the degree of marketization of urban land transfer, shown in

Table 1. Variable delimitation.

Variables	Code	Illustration
High-quality economic development	lnHED	Published by authoritative bodies
Total factor productivity	lnTFP	Solo residual method
Marketization indicator of land transfer 1	R_splot	Proportionality method, the number of land bidding, auctioning and listing cases/the number of land transfer cases
Marketization indicator of land transfer 2	r_squre	Proportional method, land bidding, auction and listing area/land transfer area
Marketization indicator of land transfer 3	r_splot_gy	Proportionality method, the number of land bidding, auctioning and listing cases/the number of land supply cases
Marketization indicator of land transfer 4	r_squre_gy	Proportionality method, land bidding, auction and listing area/land supply area
Marketization indicator of land transfer 5	r1	The weight method, by number of cases, changes the price weight
Marketization indicator of land transfer 6	r2	The weight method, by area, varies the price weights
Marketization indicator of land transfer 7	r1_g	The weighting method, which is weighted by number of cases, fixed price
Marketization indicator of land transfer 8	r2_g	Weighting method, by area, fixed price weighting
Land financing	lnrz	Sum of financing in the primary and secondary markets of land (RMB100 million)
Land primary market financing	lnrz_1	Income from land transfer fees, rents and other land supply (RMB100 million)
Secondary market financing for land	lnrz_2	Land transfer fee, land collateral and rent (100 million yuan)
Capital stock	lnk	Estimated based on the “perpetual inventory method” (billion yuan)
Number of people employed	lnlabour	Logarithmic number of people in formal and informal employment (10,000 people)
Human capital	lnhstudent	Logarithmic number of students enrolled in institutions of higher learning (10,000)
Infrastructure	lnroad	Logarithmic number of road paving area at the end of the year (10,000 square meters)
Proportion of secondary production	r_gdp2	Secondary industry value added/GDP
City area	lnarea	Logarithmic urban area (10,000 square kilometers)
Population density	lndensity	Logarithmic population density (person km2)
Fiscal gaps in the budget	lndefici	Logarithmic budgeted fiscal expenditure minus fiscal revenue (100 million yuan)
Economic openness	open	FDI as a share of fixed asset investment (%)
Total land supply	lntsland	Logarithmic of land supply area (hectares)
Proportion of land used for business purposes	r_jy	Operational land supply/total supply land area (%)
Virtual variables for the marketization reform of industrial land	policy2007	When the current year is greater than 2007, policy2007 takes 1, otherwise it takes 0

. In the empirical study, the market-oriented indicators of land transfer were used as the benchmark index for the proportion of land parcels of bidding, auctioning and listing, and other algorithmic indicators were used as the robustness test. Land financing calculations: land financing includes both income from land transfer, leasing and other supply activities of local governments in the primary market, as well as cash flow from land transfers, mortgages and leases by governments and enterprises in the secondary market. The corresponding values of the above indicators of the city were published in the “China Land and Resources Yearbook” from 1998 to 2009, and can be obtained by direct total calculation.

High-quality economic development (HED): since the 19th National Congress, high-quality economic development has become the direction of China’s future development. There are two main types of perspectives on the connotation of high-quality economic development. The first of these is to combine the “five concepts” with the main contradictions of society, the second category analyzes high-quality development from the macro, mesa and micro levels. In fact, the overall meaning of the two classification perspectives is the same, and its essential connotation is to meet the people’s growing needs for a better life as the goal of efficient, fair and green sustainable development. The Regional Development and People’s Livelihood Index is jointly released by the Statistical Society of China and the National Bureau of Statistics. The evaluation index system covers economic development, people’s livelihood improvement, social development, ecological construction, scientific and technological innovation and public evaluation along six aspects, involving 42 indicators, as a proxy variable for high-quality economic development, it not only coincides with the essential connotation of high-quality development, but also reached a certain consensus in the academic community. Therefore, this paper uses the regional development and people’s livelihood index as an alternative variable for high-quality economic development, and expands the index to 2019, so as to obtain the variable of the level of high-quality economic development, which is recorded as HED.

(2) Other indicators. This article uses the common “perpetual inventory method” to estimate the capital stock of a city. Initial capital estimates were calculated using the methods of Hall and Jones (1999), using investment data from the initial year divided by 10%. The annual depreciation rate of capital δ taken at 6% as previously documented. Then, using the annual investment data of each city after price reduction, the capital stock of each city in the calendar year was calculated. In this paper, the logarithmic value of the number of students enrolled in prefecture-level municipal institutions of higher learning (lnhstudent) is used as the proxy variable. The definition of the meaning of the variables mainly used in the empirical research in this paper is shown in Table 1, and the descriptive statistics of variables are shown in

Table 2. Summary statistics.

Variables	Obs	Mean	SD	Min	Max
lnHED	5964	4.3633	1.0241	1.6522	7.6614
lnTFP	5964	−0.6822	0.7391	−3.6406	1.0826
R_splot	5680	0.4817	0.3314	0.0000	1.0000
r_squre	5680	0.7028	0.2902	0.0000	1.0126
r_splot_gy	5680	0.3437	0.2583	0.0000	1.0000
r_squre_gy	5680	0.4459	0.2376	0.0000	1.0000
r1	5396	0.5717	0.2111	0.1685	0.9251
r2	5396	0.5216	0.2232	0.1113	0.8955
r1_g	5396	0.4891	0.1871	0.1225	0.8056
r2_g	5396	0.5286	0.1945	0.1462	0.8597
lnrz	5112	2.3121	1.8562	−5.3927	9.2878
lnrz_1	5112	2.2115	2.2773	−8.8042	7.6301
lnrz_2	5112	1.6172	2.0445	−5.9076	9.2516
lnk	5964	7.1946	1.2996	3.2168	11.2697
lnlabour	5964	4.0239	0.8253	1.7191	7.4556
lnhstudent	5964	0.8051	1.4483	−4.6053	4.6473
lnroad	5964	6.6263	1.0236	1.9456	9.9757
r_gdp2	5964	0.4759	0.1147	0.0903	0.9096
lnarea	5964	0.0976	0.8468	−4.2906	3.2327
lndensity	5964	5.7335	0.9021	1.5471	9.3555
lndefici	5964	3.6226	1.1835	−3.5066	7.5606
open	5964	0.0791	0.1426	0.0000	3.2741
lntsland	5964	6.3762	1.3051	−2.2077	11.2336
r_jy	5680	0.6273	0.2127	0.0100	0.9948
policy2007	5964	0.4715	0.4991	0.0000	1.0000

.

3.3. Model Setting

This paper first examines the impact of land transfer marketization on the high-quality development of urban economy through theoretical analysis, and then sets up a measurement model to empirically test the effect and mechanism of land transfer marketization on the high-quality development of the urban economy.

(1) The measurement model of the impact effect of land transfer marketization on high-quality economic development. In order to analyze the impact of land transfer marketization on high-quality economic development, this paper introduces the indicators of new land input and land transfer marketization in the Cobb–Douglas production function. In the input–output equation with A technical level, K capital stock, N labor input, and L new land input, the proportion of land allocated by the market is r ≤ 0 ≤ r ≤ 1. Then, the equation for the production function is:

$$Y=A{K}^{\alpha }{N}^{\beta }{(\mathrm{r}\mathrm{L})}^{\gamma }$$

(1)

Among these, $0<\alpha ,\beta ,\gamma <1$. Logarithmic to Formula (1) and controlling other influencing factors, a benchmark metrology regression model can be built:

$$Y_{it}={\alpha }_0+{\alpha }_1\ln K_{it}+{\alpha }_2\ln N_{it}+{\alpha }_3\ln L_{it}+{\alpha }_4{r}_{it}+{\alpha }_5{Z}_{it}+{\mu }_i+{\upsilon }_t+{\epsilon }_{it}$$

(2)

Among these, $\ln K$ is the capital stock, $\ln N$ is the number of people employed, $\ln L$ is the new land input, $r$ is the marketization index of land transfer. $Z_{it}$ is other control variables that have an important impact on the high-quality development of urban economies, including urban human capital, infrastructure and industrial structure.${\mu }_i$ is the urban fixation effect, ${\upsilon }_t$ is the time fixation effect, ${\epsilon }_{it}$ is a random perturbation term.

In order to prevent the bias of econometric model settings and the endogenous problem of the model, this paper constructs a dynamic model by introducing the lag terms of the dependent variables and obtains a consensus estimate of these variables [61]. From a practical point of view, there is a construction period for the new land to be put into production, and only after the completion of the project construction can the impact on economic development be fully expressed. Therefore, the degree of marketization of land transfer will affect the efficiency of land resource allocation in the current year and a long lag period, and then affect economic growth. This paper sets the project construction period to four years, so as to control the current and three-phase lag items of the land transfer marketization indicators in the measurement model. Using the methods of Arellano and Bond (1991), this paper builds the following dynamic autoregressive panel model:

$$Y_{it}={\alpha }_0+{\alpha }_1\ln K_{it}+{\alpha }_2\ln N_{it}+{\alpha }_3\ln L_{it}+{\alpha }_4{r}_{it}+{\alpha }_5{Z}_{it}+{\mu }_i+{\upsilon }_t+{\epsilon }_{it}$$

(3)

Among these, $\ln Y_{i,t-1}$ is a first-order laggard term for high-quality economic development, and $r_{it}$ is the current period and the third phase lag item of the indicator of the degree of marketization of land transfer. In order to observe the impact of the market-oriented reform of industrial land transfer in 2007, the above formula also adds the dummy variable policy2007 of the market-oriented reform of industrial land transfer, and the interactive term policy2007 ×r of policy2007 and the marketization index of land transfer.

(2) The impact mechanism of land transfer marketization on high-quality economic development was tested. The first is the impact of the marketization of land transfer on land financing (lnrz). The level of land financing is based on the market value of the land, and the market value of the land is determined by the market price of the current period, so it is mainly the market-oriented level of land transfer in the current period that affects land financing. Therefore, only the impact of land marketization indicators on land financing in the current period was analyzed. This article constructs the following dynamic autoregressive model:

$$Y_{it}={\alpha }_0+{\alpha }_1\ln K_{it}+{\alpha }_2\ln N_{it}+{\alpha }_3\ln L_{it}+{\alpha }_4{r}_{it}+{\alpha }_5{Z}_{it}+{\mu }_i+{\upsilon }_t+{\epsilon }_{it}$$

(4)

Among these, ${\mathrm{lnrz}}_{it-1}$ is a delay in the first phase of land financing. The control variables $Z_{it}$ include the city’s industrial structure, land scarcity, population density, and budgetary fiscal gaps.

The second is the impact of land transfer marketization on the efficiency of urban resource allocation. This paper uses total factor productivity (TFP) to measure resource allocation efficiency. Due to the construction cycle of the project, the marketization indicator of the lag period will affect the TFP. Therefore, the following dynamic autoregressive model was constructed to empirically study the influence of land transfer marketization on urban TFP, and the independent variables included the current and three-phase lag items of land transfer marketization indicators.

$$Y_{it}={\alpha }_0+{\alpha }_1\ln K_{it}+{\alpha }_2\ln N_{it}+{\alpha }_3\ln L_{it}+{\alpha }_4{r}_{it}+{\alpha }_5{Z}_{it}+{\mu }_i+{\upsilon }_t+{\epsilon }_{it}$$

(5)

According to the relevant literature, this paper also controls the capital stock, infrastructure, human capital, economic openness of the city in the regression model, and also controls the total land supply. In addition, in order to analyze the relationship between the land financing effect and the resource allocation effect, a land financing variable was added to the model of Equation (5).

Since the core explanatory variable land transfer marketization index has a lag item, the measurement method in this paper adopts a two-step differential GMM estimate. As a robustness test, a panel fixed effect estimation and two-step system GMM estimation were also used in this paper. The exogenous tool variables use the size of the city in 1996 and the shortest distance from the city to the three major ports of Tianjin, Shanghai, and Hong Kong, and all interact with the corresponding year.

4. Empirical Results

4.1. Preliminary Regression Results

According to the econometric model Formulas (2) and (3), after controlling the variables of capital, labor, human capital, infrastructure, industrial structure, land supply area, and industrial land transfer marketization reform at the city level, the impact of the degree of land transfer marketization on urban lnHED was estimated based on the urban panel data from 1999 to 2019. The indicators of the degree of marketization of land transfer in

Table 3. The impact of land transfer marketization on high-quality economic development.

Interpreted Variables lnHED	(1)	(2)	(3)	(4)	(5)	(6)
	FE		Differential GMM		System GMM
r_splot	0.0318 *** (0.0067)	0.0558 *** (0.0086)	−0.0037 *** (0.0008)	0.0072 *** (0.0013)	−0.0213 *** (0.0012)	−0.0149 *** (0.0014)
l.r_splot	0.0129 * (0.0073)	0.0162 ** (0.0075)	0.0042 *** (0.0011)	0.0035 *** (0.0008)	0.0063 *** (0.0009)	0.0063 *** (0.0011)
l2.r_splot	0.0057 (0.0073)	0.0081 (0.0073)	0.0016 ** (0.0007)	0.0012 ** (0.0006)	0.0031 *** (0.0008)	0.0023 ** (0.0009)
l3.r_splot	0.0008 (0.0065)	0.0039 (0.0065)	0.0027 *** (0.0007)	0.0034 *** (0.0006)	0.0019 *** (0.0008)	0.0013 ** (0.0006)
policy2007		0.5618 *** (0.0092)		0.1295 *** (0.0021)		0.0246 *** (0.0013)
policy2007×r_splot		−0.0481 *** (0.0106)		−0.0205 *** (0.0012)		−0.0122 *** (0.0021)
control variables	YES	YES	YES	YES	YES	YES
City FE			YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	5565	5565	5376	5376	5376	5376
Cities	265	265	256	256	256	256
Adj R2	0.7366	0.7369
AR1			0.0000	0.0000	0.0000	0.0000
AR2			0.4986	0.4675	0.4547	0.4536
Sargan P			1.0000	1.0000	1.0000	1.0000

Note: GMM estimates are two-phase estimates. *, ** and *** represent significant levels of 1%, 5%, and 10%, respectively, with bracket values being standard errors.

are r_splot, and the included models have the values of current period and lagging three periods. The autocorrelation test for GMM estimated residuals is neither significant nor the Sargan P-value of the tool over-recognition test, indicating that there is no second-order series correlation between the residuals and explanatory variables, and that the tool variable selection in GMM regression is appropriate and valid. The GMM estimate is thus reasonably valid. Comparing the column (4) differential GMM and column (6) system GMM estimates, the estimation results of the column (4) two-stage differential GMM on the key explanatory variables and control variables are closer to the regression results of the fixed effects of the column (2) panel. Therefore, the results of the differential GMM estimates were mainly used in this paper.

According to column (3) of Table 3, the market-oriented indicators of land transfer are significantly negative in the current period and the lag period is significantly positive. In column (4), the marketization indicators of land transfer are significantly positive in both the current period and the lag period, and the virtual variable policy2007 of the marketization reform of industrial land transfer is also significantly positive; however, the interaction between policy2007 and the current indicators of land transfer marketization is significantly negative. The results of the return before 2008, regardless of the impact of the interaction item, clearly show that the marketization indicators of land transfer have a positive impact on high-quality economic development in both the current period and the lag period. After 2008, considering the impact of the interaction term, the coefficient of the r_splot of the current land transfer marketization index in the current period is −0.0134 (0.0071–0.0205) in the calculation of the differential GMM estimate in column (4), and the marginal effect of the increase in the marketization of land transfer on the high-quality economic development in the short term is negative. However, if we take into account the r_splot of the market-oriented index of land transfer after 2008, the average value is 0.600, the net effect of the market-oriented transfer reform of industrial land on the high-quality economic development is 0.1172 (0.1295 ‒ 0.0205 × 0.6000), and the impact of the market-oriented reform of industrial land has improved the quality of economic development at one time.

To sum up, before 2008, the marketization of land transfer promoted high-quality economic development in both the short and long term. After the market-oriented transfer reform of industrial land in 2008, the marginal effect of the market-oriented indicators of land transfer decreased, and the impact on high-quality economic development in the short term was negative, but the direct effect of the market-oriented reform of industrial land transfer on the high-quality development of the urban economy was positive. That is to say, the results of the return show that the transfer of industrial land has changed from an agreement to a more market-oriented bidding, auction and listing method, and in the short term, the economic growth rate has been reduced due to the loss of investment projects promoted by some local governments at lower land prices, and the reform of market-oriented transfer of industrial land has indeed impacted the local government’s model of “seeking development with land” in the short term. However, in the long run, the market-oriented allocation of land may improve the efficiency of resource allocation and promote high-quality economic development by attracting better investment projects.

4.2. Mechanism Analysis

According to the previous research assumptions and econometric models, this paper further examines the impact mechanism of land transfer marketization on high-quality economic development from two aspects:

(1) The financing effect of land transfer marketization was tested.

Table 4. The impact of the marketization of land transfer on land financing.

Interpreted Variables Land Financing	(1)	(2)	(3)	(4)	(5)	(6)
	FE			Differential GMM
	lnrz	lnrz_1	lnjrdk	lnrz	lnrz_1	lnrz_2
r_splot	0.2447 * (0.1412)	0.5690 *** (0.0645)	0.0895 *** (0.0207)	0.7706 *** (0.0086)	0.8482 *** (0.0065)	0.4269 *** (0.0292)
control variables	YES	YES	YES	YES	YES	YES
City FE				YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	5922	5985	5964	5376	5460	5187
Cities	282	285	284	256	260	247
Adj R2	0.3266	0.7571	0.6933
AR1				0.0000	0.0000	0.0000
AR2				0.9772	0.8136	0.2215
Sargan P				0.9596	0.9913	0.9998

Note: For comparison purposes, all sample estimates are from 1999 to 2008. (1) and (4), the explanatory variables are the land financing of the city; (3), the loan balance of the financial institutions at the end of the year in the city; (2) and (5), the financing of the primary market of land; and (6), the financing of the secondary market of land. * and *** represent significant levels of 1% and 10%, respectively. The values within the parentheses are standard errors.

shows the estimated effects of land transfer marketization on land financing under different model settings. Regression controls population density, urban area, fiscal gap within the government budget, proportion of secondary industry, year effect and regional effect. From the results of the return, it can be seen that the market-oriented indicators of land transfer have a significant positive impact on land financing, and the impact on the primary market of land is significantly greater than that of the secondary market. According to the return results of column (4), under the condition that other conditions remain unchanged, the level of urban land transfer marketization increased by one percentage point, and the overall land financing level of the city increased by 0.7700% in that year. Column (3) of Table 4 uses the loan balance of financial institutions at the end of the year in cities as an alternative explanatory variable for land financing, and the results are also significantly positive: the level of marketization in cities increased by one percentage point, and the balance of loans of financial institutions in cities at the end of the year increased by 0.0900%. Therefore, the marketization of urban land transfer has a significant financing effect by increasing the degree of “land capitalization”. It has been proposed that the expansion of the scale of land financing in an urban area can increase urban construction financing and reduce the cost of urban financing, while the expansion of urban credit can promote high-quality economic development by increasing investment [62]. In order to confirm this mechanism, this paper uses the model of Table 4 to take urban investment as the explanatory variable, and empirically tests whether the marketization of land transfer promotes the increase of urban investment through the expansion of land financing scale, and the results are shown in

Table 5. The impact of land financing on investment.

Interpreted Variables lninvest	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	FE			Differential GMM
lnrz	0.0855 *** (0.0071)			0.0908 *** (0.0005)
lnrz_1		0.1367 *** (0.0056)			0.0834 *** (0.0012)
lnrz_2						0.0340 *** (0.0002)
r_splot	0.2479 *** (0.0381)	0.0720 *** (0.0643)	0.1498 *** (0.0217)	0.1732 *** (0.0016)	0.0608 *** (0.0067)	0.2112 *** (0.0009)	0.1712 *** (0.0046)
control variables	YES	YES	YES	YES	YES	YES	YES
City FE				YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES
N	5922	5985	5985	5376	5460	5187	5187
Cities	282	285	285	256	260	247	247
Adj R2	0.8573	0.9456	0.9373
AR1				0.0000	0.0000	0.0001	0.0000
AR2				0.0192	0.9566	0.0263	0.8131
Sargan P				0.9817	1.0000	0.9956	1.0000
Sobel Z	1.7160	8.3250	-	82.5155	65.8102	14.6565	-

Note: *** represent significant levels of 10%. The values within the parentheses are standard errors.

. Different regression models show that the expansion of land financing scale has a significant role in promoting urban investment. It can be seen from the regression results of column 5 (7) of Table 5 that the level of land transfer marketization increased by one percentage point, and urban investment increased by 0.17%. According to Baron and Kenny (1986) and Wen and Ye (2014), the intermediate effect of land financing is examined by step-by-step regression coefficients [60,61], and the results of Table 5 show that the mediating effect of land financing is significant. Combined with Table 4, differential GMM estimates that the proportion of the financing intermediary effect in the primary land market to the total effect is 41.3200% (0.8482×0.0834/0.1712), while the intermediary effect of financing in the secondary land market accounts for only 8.4800% of the total effect. Therefore, local governments have a greater effect on stimulating investment growth through primary land market financing.

(2) Test of the effect of resource allocation in the marketization of land transfer. Another channel for the marketization of land transfer to affect economic growth is to change the efficiency of resource allocation, which is reflected in the impact on urban TFPs. The resource allocation effect of the land transfer marketization index was tested under different model settings, and the results are shown in

Table 6. Estimation of the market-oriented resource allocation effect of land transfer.

Interpreted Variables lnTFP	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	1999–2008			1999–2019
r_splot	0.0078 ** (0.0011)	0.0023 ** (0.0012)	0.0177 ** (0.0008)	0.0145 ** (0.0013)	0.0084 ** (0.0017)	0.0033 * (0.0018)	0.0117 ** (0.0018)	0.0347 ** (0.0031)
l.r_splot	0.0223 ** (0.0011)	0.0172 ** (0.0013)	0.0182 ** (0.0007)	0.0235 ** (0.0016)	0.0272 ** (0.0017)	0.0198 ** (0.0022)	0.0203 ** (0.0018)	0.0391 ** (0.0026)
l2.r_splot	0.0271 ** (0.0007)	0.0229 ** (0.0007)	0.0238 ** (0.0009)	0.0163 ** (0.0010)	0.0174 ** (0.0011)	0.0135 ** (0.0013)	0.0131 ** (0.0013)	0.0026 ** (0.0014)
l3.r_splot	0.0136 ** (0.0005)	0.0106 ** (0.0005)	0.0086 ** (0.0006)	0.0076 ** (0.0009)	0.0091 ** (0.0008)	0.0066 ** (0.0009)	0.0072 ** (0.0007)	0.0018 ** (0.0011)
policy2007					0.0160 ** (0.0022)	0.0103 ** (0.0025)	0.0173 ** (0.0021)
policy2007×r_splot					0.1039 ** (0.0082)	0.1071 ** (0.0092)	0.0718 ** (0.0048)
Logarithm of R&D personnel		−0.0124 *** (0.0017)				0.0045 ** (0.0021)
R&D investment intensity			−0.0437 ** (0.0019)				−0.0135 ** (0.0027)
Average R&D intensity of listed companies								0.0021 ** (0.0004)
control variables	YES	YES	YES	YES	YES	YES	YES	YES
City FE				YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	5250	5250	5250	5334	5334	5334	5334	4494
Cities	250	250	250	254	254	254	254	214
AR1	0.0046	0.0029	0.0028	0.0000	0.0000	0.0001	0.0000	0.0001
AR2	0.1426	0.1655	0.1791	0.0532	0.0567	0.0558	0.0636	0.0325
Sargan P	0.9177	0.9955	0.9935	0.9817	1.0000	0.9956	1.0000	1.0000

Note: The estimation method is a two-phase differential GMM estimate. *, ** and *** represent significant levels of 1%, 5%, and 10%, respectively. The values within the parentheses are standard errors.

. From column (1) to (8), the current and delayed land transfer marketization indicators have a significant positive impact on urban TFPs. Column (5)–column (7) show that the virtual variable index of industrial land marketization reform policy2007 is significantly positive, indicating that the direct effect of market-oriented transfer of industrial land on resource allocation is significantly positive. The interaction between the marketization reform of industrial land and the land marketization index is also significantly positive, indicating that the marketization reform of industrial land has enhanced the resource allocation effect of the land transfer marketization index after 2008.

Since TFP is a comprehensive solo residual value that includes both the resource allocation effect and the technological progress effect, this paper adds a technical input control variable to the model to eliminate the technical effect of TFP to better analyze the resource allocation effect of land transfer marketization. In columns (2), (3), (6), (7) and (8) of Table 6, the logarithm of R&D personnel and the intensity of R&D investment (due to the lack of data in prefecture-level cities, this is provincial-level data), as well as the average R&D intensity of listed companies in prefecture-level cities, are controlled successively. According to the results of Table 6, after controlling different technical input indicators, the impact of land transfer marketization indicators on urban resource allocation in both the current period and the lag period is still significantly positive. Therefore, the marketization of land transfer has improved the efficiency of resource allocation in the city, which is manifested in the improvement of TFP. In addition, the market-oriented reform of industrial land transfer has also improved the allocation efficiency.

(3) The impact of land financing on the effect of resource allocation. In order to analyze the impact of land financing on resource allocation, after controlling a series of factors affecting urban resource allocation, land financing was added as an explanatory variable. The explanatory variables were measured by different indicators of total urban land financing (lnrz), land primary market financing (lnrz_1), land secondary market financing (lnrz_2) and loan balance of financial institutions at the end of the year (lnjrdk). In addition, land financing may also have a lagging impact on resource allocation, so the results of the lag model are also reported, as shown in

Table 7. The impact of land financing effects on resource allocation effects.

Interpreted Variables lnTFP	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	1999–2008							2009–2019
	lnrz		lnrz_1		lnrz_2		lnjrdk	lnrz_1	lnjrdk
Current land financing	−0.0041 *** (0.0004)	−0.0016 *** (0.0002)	−0.0021 *** (0.0002)	−0.0013 *** (0.0001)	0.0019 *** (0.0001)	−0.0001 (0.0003)	−0.0706 *** (0.0013)	0.0082 *** (0.0004)	−0.0014 (0.0016)
The first phase of land financing is lagging behind		0.0022 *** (0.0004)		−0.0013 *** (0.0001)		0.0007 ** (0.0003)	0.0363 *** (0.0011)	0.0004 (0.0003)	0.0133 *** (0.0013)
The second phase of land financing is lagging behind		0.0030 *** (0.0005)		−0.0020 *** (0.0002)		0.0003 ** (0.0002)	0.0166 *** (0.0012)	−0.0085 *** (0.0002)	0.0263 *** (0.0010)
control variables	YES	YES	YES	YES	YES	YES	YES	YES	YES
City FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
N	5229	5061	5250	5250	5208	4347	5208	5292	5292
Cities	249	241	250	250	248	207	248	252	252
AR1	0.0069	0.0373	0.0050	0.0053	0.0761	0.0732	0.0067	0.0000	0.0000
AR2	0.1201	0.2038	0.1644	0.1439	0.1254	0.8034	0.4455	0.1386	0.1498
Sargan P	0.9911	0.9995	0.9970	0.9920	0.9980	1.0000	1.0000	1.0000	1.0000

Note: The estimation method is a two-phase differential GMM estimate. The explanatory variables are lnrz, lnrz_1, lnrz_2, and lnjrdk. ** and *** represent significant levels of 5%, and 10%, respectively. The values within the parentheses are standard errors.

. According to columns (1) and (2) of Table 7, the TFP impact of land financing on cities in the current period is significantly negative, but the impact is positive in the lag period. Column (7) uses the year-end financial institution loan balance as a substitute variable result. Therefore, the impact of land financing on resource allocation is more complex, which has both a crowding out effect and a promoting effect. Subdividing the impact of land primary and secondary market financing, it is found that the impact of land primary and secondary market financing in 1999–2008 is diametrically opposite, and the impact of land primary market financing on resource allocation is negative regardless of the current period or lag period, while the impact of land financing in the secondary market is only observed as a significant positive impact. In addition, the impact is different from time to time. For the 2009–2019 sample, column (8) shows that the impact of primary market financing on land also observed a significant positive impact in the current period, and column (9) shows that the negative impact of the financing index in the current period was no longer significant. The lag period and total impact were significantly positive. In the primary land market, local governments supply land as monopolies to obtain land transfer income, while in the secondary land market, whether funds are raised or used, the degree of marketization is higher. These results reflect a misallocation of resources due to improper government intervention. If the improper intervention of the government is reduced and the role of the market is expanded, then the market will be conducive to a reduction of the crowding out effect of land financing on resource allocation and will play a more promoting role. After 2008, the positive effect of financing on resource allocation in the primary land market may be the positive effect of further marketization of land transfer.

(4) The intermediate variable method test of the influencing mechanism. In order to further analyze the impact mechanism of land transfer marketization on economic growth and explain the impact of land financing and resource allocation on economic growth, this paper uses a benchmark measurement model to test the intermediary effect of land transfer marketization, and the results are shown in

Table 8. The intermediate variable method test of the impact mechanism of land transfer marketization.

Interpreted Variables lnHED	(1)	(2)	(3)	(4)	(5)	(6)	(7)
r_splot	0.0270 *** (0.0053)		0.0238 *** (0.0052)	0.0039 (0.0048)		0.0057 (0.0058)
lnrz_1		0.0140 *** (0.0021)	0.0133 *** (0.0025)	0.0084 *** (0.0017)	0.0085 *** (0.0017)
lnTFP				0.2004 *** (0.0065)	0.2011 *** (0.0064)	0.2032 *** (0.0065)	0.2043 *** (0.0064)
control variables	YES	YES	YES	YES	YES	YES	YES
City FE	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES
N	5565	5565	5565	5565	5565	5565	5565
Cities	265	265	265	265	265	265	265
Adj R2	0.8611	0.8022	0.8164	0.7988	0.7633	0.8071	0.8210
Sobel Z	-	-	4.8222	-	21.0481	20.8822	-

Note: The estimation method is a fixed-effect estimate. *** represent significant levels of 10%. The values within the parentheses are standard errors.

. According to Baron and Kenny (1986) and Wen and Ye (2014), the mediation effect was tested by step-by-step regression coefficients [63,64], regression was directly estimated using panel fixed effects, and the land transfer marketization indicator uses only the current value. According to the previous empirical results, the marketization of land transfer has a greater impact on the financing of the primary market of land, and the impact of financing in the primary market of land on the effect of resource allocation is more complex, so the land financing index adopts the financing of the primary market of land. According to MacKinnon et al. (1995) [65], the Sobel Z values of columns (3) and (6) in Table 8 show that the impact of land transfer marketization on economic growth, land financing and the intermediary effect of urban TFP are significant. Land financing based on regression coefficients and significant has a partial mediating effect, while TFP has a fully mediating effect. In addition, in column (5), regarding the impact of land financing on economic growth, the mediating effect of TFP is also significant, and there are some intermediary effects combined with regression coefficient and significance. Further, according to the regression coefficient and significance of column (4), the impact of land transfer marketization on economic growth, land financing and urban TFP have a complete intermediary effect.

The above results further illustrate that the marketization of land transfer has promoted high-quality economic development through financing effects and resource allocation effects. On the one hand, land financing promotes high-quality economic development through direct intermediary effects, and on the other hand, it also affects high-quality economic development by influencing resource allocation.

4.3. Endogenous Treatment and Robustness Testing

There may be an endogenous nature between the way land is transferred and economic growth. Cities with better economic development may have more financial resources and willingness to accelerate the process of land transfer marketization. However, the main model in this paper is a two-stage GMM estimate that can alleviate endogenous concerns. In addition, by introducing the three-phase lag item of the degree of marketization of land transfer, the endogenous problem between the degree of marketization of land transfer and the high-quality development of the urban economy can also be further alleviated.

The impact effect of land transfer marketization on high-quality economic development may be affected by market-oriented measurement indicators.

Table 9. Economic high-quality effect estimation for different market-oriented indicators.

Interpreted Variables lnHED	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	r_splot	r_squre	r_splot_gy	r_squre_gy	r1	r2	r1_g	r2_g
r_	−0.0036 *** (0.0008)	0.0125 *** (0.0007)	0.0010 (0.0008)	0.0056 *** (0.0007)	−0.0001 (0.0007)	0.0018 ** (0.0010)	0.0205 *** (0.0017)	0.0038 *** (0.0008)
l.r_	0.0041 *** (0.0011)	0.0031 *** (0.0004)	0.0009 (0.0007)	−0.0009 ** (0.0006)	−0.0030 *** (0.0006)	−0.0006 (0.0005)	−0.0054 *** (0.0012)	−0.0022 *** (0.0006)
l2.r_	0.0015 ** (0.0008)	0.0049 *** (0.0003)	−0.0010 (0.0005)	0.0010 *** (0.0004)	0.0021 *** (0.0006)	0.0017 *** (0.0004)	0.0002 (0.0006)	−0.0007 (0.0006)
l3.r_	0.0026 *** (0.0005)	0.0064 *** (0.0006)	0.0022 *** (0.0008)	0.0032 *** (0.0005)	0.0035 *** (0.0005)	0.0027 *** (0.0005)	0.0007 (0.0007)	0.0022 *** (0.0006)
control variables	YES	YES	YES	YES	YES	YES	YES	YES
City FE	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	5376	5376	5376	5376	5376	5376	5376	5376
Cities	256	256	256	256	256	256	256	256
AR1	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000	0.0000
AR2	0.4977	0.2215	0.4976	0.1878	0.2276	0.2034	0.4113	0.1927
Sargan P	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000

Note: The estimation method is a two−phase differential GMM estimate. ** and *** represent significant levels of 5%, and 10%, respectively. The values within the parentheses are standard errors.

presents the two-stage differential GMM estimates for different land transfer marketization indicators. The current impact of the different indicators in Table 9 is positive or negative, or not significant, but the lag period shows a positive impact and is basically significant. These return results further confirm that the market-oriented transfer of land has a significant role in promoting the high-quality development of the urban economy at the lag time.

The market-oriented reform of land transfer mainly involves commercial land. So, will the proportion of operational land in the total land supply of the city have an impact on the results of the marketization of land transfer? The regression results of this paper after controlling the proportion of operational land are similar to the basic conclusions in Table 9. This paper also attempts to use different marketization indicators to examine the impact of land transfer marketization on urban TFPs. All regression results are also basically the same as those reported in this article. The relevant results of these returns are limited in space and are not reported, but are available on request.

5. Conclusions and Policy Implications

5.1. Conclusions

The market-oriented reform of urban land transfer is a major institutional change with a far-reaching impact on China’s economic development, but the existing literature is insufficient to study the economic impact of this important system. This paper uses panel data from 1999 to 2019 in China’s prefecture-level and above cities to study the impact mechanism and influence effect of the degree of marketization of land transfer on economic growth. Empirical studies of multiple metrics based on multiple model settings, multiple estimation methods and key variables have found that the marketization of land transfer has significantly promoted urban economic growth in the long run. Before 2008, the marketization of land transfer promoted high-quality economic development in both the short and long term. After the reform of market-oriented transfer of industrial land in 2008, the marginal effect of the market-oriented indicators of land transfer declined, and the impact on high-quality economic development in the short term was negative, but the direct effect of the market-oriented reform of industrial land transfer on the high-quality development of the urban economy was positive. Moreover, not only commercial and residential land, but also the improvement of the level of marketization of industrial land transfer has also had a positive impact on economic growth. Specific to the impact mechanism, this paper finds that the marketization of land transfer will affect economic growth through financing effects and resource allocation effects. The measurement evidence shows that the marketization of land transfer has a significant financing effect, and the impact of land financing on economic growth is significantly positive. At the same time, the marketization of land transfer has a significant resource allocation effect, which is manifested in improving the TFP of the city. At the same time, this paper also finds that land financing has a very complex impact on the effect of resource allocation, which is embodied in the crowding out effect of financing in the primary market of land on TFP, and the impact of financing in the secondary market of land on TFP is to promote it.

5.2. Policy Recommendations

The research conclusions of this paper provide an explanation, based on the perspective of land resource allocation systems, to the mystery of China’s high-quality economic development, enrich the relevant research on land resource allocation theory, and also provide a preliminary assessment of the economic growth performance of China’s land system reform, which has important policy implications for deepening the reform of the land use system.

(1) Continue to improve the land transfer system in the direction of marketization. From the empirical results, it can be seen that the increase in the degree of marketization of land transfer is conducive to the long-term growth of the economy and the improvement of the efficiency of resource allocation. Many people in the government and academia have worried that excessive marketization of land will increase the cost of urban operation, especially in squeezing out industrial projects. The research in this paper shows that the marketization of land transfer does have a negative effect in the short term, but that the overall effect on economic growth is significantly positive in the long run. Specific to the transfer of industrial land, from an agreement to a more market-oriented way of bidding, auctioning and listing, there may be, in the short term, a reduction in the economic growth rate due to the loss of investment projects promoted by some local governments at lower land prices. However, in the long run, the market-oriented allocation of land can promote the long-term sustainable growth of the economy by attracting better investment projects and improving the efficiency of resource allocation. Therefore, land transfer and its related systems should be firmly improved in the direction of marketization, rather than taking the road of anti-marketization. At the same time, it is necessary to define property rights in accordance with the laws of the market economy, thereby reducing the friction of interests, improving the efficiency of resource allocation, and making the allocation of resources reach an optimal Pareto state. It is also necessary to give full play to the role of entrepreneurship in the allocation of resources, further improve the policies and legal systems for protecting and motivating enterprises and entrepreneurs, accelerate the establishment of a modern corporate governance system, and make every effort to create a social and cultural atmosphere that supports the growth of entrepreneurs.

(2) Improve the construction of the primary and secondary land market system and prevent land financing risks. Government supervision and intervention should make up for market failures, promote the positive effects of marketization, and weaken the negative effects of marketization. This paper finds that land financing in the primary market will generally hinder the allocation of resources. This phenomenon is likely to indicate the inefficient use of land transfer income funds by local governments, but also shows that local governments create a great distortion in behavior the primary land market. It is therefore urgent to prevent risk associated with local government land financing, reduce local government distortions as much as possible in the primary land market, and give more play to the role of market mechanisms. At the same time, it is necessary to strengthen the effective supervision of local government land transfer revenue funds and improve the efficiency of capital utilization. The empirical evidence also shows that the financing of the secondary market for land generally promotes the allocation of resources, so the construction of the secondary market for land should be vigorously strengthened, a diversified land supply should be encouraged, the land leasing and transfer market activated, and suitable land financing models for local governments and enterprises should be actively explored.

(3) Transform the functions of local governments, promote the reform of the local fiscal and taxation system, and accelerate the transformation and upgrading of the economic structure. Under the local development-oriented government system, coupled with the mismatch between the rights and responsibilities of local governments and taxes, local governments are highly dependent on land transfer income and land financing. Under such conditions, local governments may hinder land transfer and the market allocation of land financing funds, thereby reducing the potential economic growth rate and resource allocation level. Therefore, the transformation of local governments into service-oriented governments should be accelerated, the reform of local fiscal and taxation systems, such as the timely introduction of real estate taxes, should be promoted, and the construction of local sustainable tax sources should replace local governments’ dependence on land transfer revenue and land financing. This will reduce the improper intervention of local governments in land supply and land financing, which is conducive to more effectively playing the role of the market in allocating resources and resolving land financing risks. Based on the scientific analysis of national conditions, we will vigorously promote the transformation of the economic structure, of which the development of a digital economy and innovation of digital technology are an important part. This will help to continuously improve digital infrastructure, promote new tools such as digital currencies and mobile payments, enter new markets, improve the internal business environment, increase residents’ income, and gradually achieve common prosperity.

5.3. Research Deficiencies and Prospects

The research in this paper is as follows. First, the “China Urban Statistical Yearbook” has been updated to 2021, but the statistical caliber of some data have not maintained continuity, so the level of high-quality development of urban economy can only be evaluated from 1999 to 2019 and cannot reflect the latest achievements. Second, the research object of this paper is a prefecture-level city in China, and it has not been extended to the county level in China, so the richness and fullness of the study are lacking. Thirdly, when performing mechanism analysis, this paper only found two mechanisms of action, namely the financing effect and the resource allocation effect, but there may be more intrinsic mechanisms in subsequent studies. Finally, based on the purpose of the research and data considerations, this paper mainly considers the behavior of local governments at the city level, so there is not much attention paid to the enterprise level, the role of entrepreneurs as the driving force of the market economy, and the neglect of capital accumulation, which is the direction that future research should seek to focus on and improve.