Abstract
In this paper, we adopt stochastic frontier model of translog production function to measure the total factor productivity growth rate of 13 cities in Jiangsu province from the year 2000 to 2020 and study the characteristics of the contribution rates of capital, labor, and total factor productivity in driving economic growth. The study shows that, firstly, the explanatory power of the model could be increased by 35.9% after 6 effect variables, such as urban economic development level, have been taken into consideration when two-step method has been adopted to explore the model parameters of stochastic frontier; secondly, there is an obviously positive correlation between the 4 variables, urban economic development level, urban size, level of opening up to the outside world and infrastructure construction, and production efficiency, while there is a negative correlation between economic structure and production efficiency, and the research input intensity has an uncertain effect on production efficiency; thirdly, the TFP growth trend of urban economy of the province shows a similar “W” trend, technical progress becomes the first element in affecting total factor productivity, and the negative growth in technical efficiency and scale efficiency result in a dramatic decline in growth rate of TFP; fourthly, among all the elements contributing to urban economic growth, capital has the greatest contribution, TFP takes the second place, and labor element takes the last. The double difference between capital and TFP appears to be the major cause in forming regional difference of urban economy. The regional differences of TFP concerning infrastructure advantage, position advantage, factor advantage, and policy advantage are the key factors in urban economic gap. At the end of this study, we propose some policy suggestions.
1. Introduction
In neoclassical economics, the increase in production factor input and technical progress are elements that can affect economic growth [1]. In the background of resource constrain, the mode of economic development that inputting factors increase economic growth cannot be kept on because of marginal decline of factors [2]. The fact that economic growth of China has declined since the year 2012 proves that the factor-driving mode cannot realize economic sustainable growth [3]. Total factor productivity (TFP) is an important tool for finding source of economic growth and is a major indicator in evaluating economic growth driving to force structure. Yu and Li [4] also pointed that promoting urban economic TFP has become an important direction when considering transforming economic growth driving to force structure. The relevant literature at present mainly concerns the measure and constitution of TFP, influential factors, and spatial differences. There are prominent differences in measurement and decomposition of TFP due to study methods and data time, and there is no convergence concerning TFP influential degree due to differences in factor choice and clustering method of TFP [5, 6]. A deep analysis on spatial characteristics of driving force of economic growth is conductive for realizing the reasons for the gap in regional economic growth and for clarifying whether the gap in economic development is more due to the differences in TFP or the accumulation of production factors.
At present, scholars in foreign countries use a wide range of TFP analysis and have conducted in-depth research on the impact of economic growth and obtained enlightening conclusions. Many valuable conclusions have been drawn from the research methods, research perspectives, and research paradigms. In terms of research methods, scholars used different methods to measure TFP and analyzed the determinants of growth rate. For instance, in his book, Green Productivity: Applications in Malaysia’s Manufacturing, Ahmed [7] used a method to calculate the actual total factor productivity per unit of labor growth by internalizing the pollutant emissions other than the inputs in the traditional production function. Jreisat et al. [8] used DEA based on nonparametric method to evaluate and review the productivity changes of 14 banks operating in the Egyptian market from 1997 to 2013. It is found that the Egyptian Banking Industry as a whole shows a decline in productivity, which is mainly due to technological progress. Based on the unbalanced panel data of 1994, 2004, and 2014, Mariyono [9] used stochastic frontier transcendental logarithmic production function to estimate TFP of Indian rice production. He found that the main sources of TFP growth of rice production were technological change and allocation efficiency effect, and the contribution rate of technical efficiency was low. Moreover, Mariyono [10] evaluated the implications of and assessed the economic and sustainable impact of environmentally friendly technological packages introduced into agribusiness players in the centre of chili-producing regions of Indonesia and found that the packages of technologies improved economic and sustainability performance of agribusiness resulting from efficient use of agrochemicals and water resource and increased the production. Dhandhalya and Tarpara [11] analyzed the return on investment and the source of productivity of Gujarat peanut researched from 1990 to 1991 and from 2011 to 2012 and constructed the total output, total input, and TFP index by using the Tornqvist-Theil index technology and using two outputs and ten inputs. In terms of research perspectives, scholars analyzed the growth of TFP from different perspectives. Fuglie and Wang [12] compared the growth of agricultural TFP in developed countries, developing countries, and transition countries from a national perspective and found that the overall growth of agricultural TFP in developed countries was strong. The developing countries represented by China maintained extremely high TFP growth of the past from 1991 to 2009, and TFP growth in coastal areas was significantly higher than that in inland areas. Ahmed [13] estimated TFP index of tourism industry and divided it into three parts: input change rate, output change rate, and frontier change. The author provided the efficiency measurement of 101 tourism destinations. Afzal [14] focused on estimating total factor productivity growth in Pakistan’s economy and particularly in the agriculture sector, covered total factor productivity growth period from 1973 to 2020, and found that TFP has been the lowest in the agriculture sector as compared to other sectors as per findings of the study. Malik [15] examined the output and total factor productivity (TFP) growth by estimating a stochastic frontier production function for a panel of Middle East and North Africa (MENA) countries and to decompose TFP growth into technical change, technical efficiency, and scale efficiency. In terms of research paradigms, Benzaquen [16] focused on the impact of manufacturing industry (shipbuilding industry) on economic growth. He took Peru’s large shipbuilding industry from 1969 to 1990 as an example, obtained TFP of an industry through quantitative empirical analysis, and further studied that the evolution of Peru’s shipbuilding activities coincided with the growth and decline period of Peru’s GDP. Ivanic and Martin [17] discussed the impact of factor productivity on global poverty across industries (agriculture, industry, and services) and found that the improvement on agricultural productivity usually has a greater poverty reduction effect than that of industry or services in poor countries. Wang and Xu [18] studied the direct and spatial impacts of China’s human capital structure on regional GTFP and found that the different levels of human capital have different effects on GTFP and that some types of human capital (such as tertiary-level educational human capital) promote local GTFP, while other types (such as primary-level educational human capital) inhibit it.
Among all economic plates in Chinese regional economy, the annual average economic growth rate of Jiangsu province sustains more or less 10%, and Jiangsu province is the second strongest economy in China, developing a “Jiangsu Model” in economic and social development. Nevertheless, according to the literature at home and abroad, there are few literature pieces to calculate the contribution rate of capital, labor, and TFP in economic growth and take measures to improve the contribution rate of TFP. Further, relevant literature focusing on TFP of Jiangsu cities is scarce. This paper will fill the research gap in the field of urban TFP in Jiangsu province and explore the mechanism of influencing factors on production efficiency and evolution trend of the urban TFP growth rate in the year between 2000 and 2020 of urban economy in Jiangsu province. In order to enhance the driving force of economic growth, each part of this paper is arranged as follows: the first part is to test the applicability of stochastic frontier model and give the specific form. The second is to estimate and analyze the production efficiency of 13 cities in Jiangsu. (There are 13 cities in Jiangsu province, which is divided into three regions: Southern Jiangsu, Central Jiangsu, and Northern Jiangsu. Southern Jiangsu includes five cities such as Nanjing, Suzhou, Wuxi, Changzhou, and Zhenjiang, Central Jiangsu includes three cities such as Nantong, Taizhou, and Yangzhou, and Northern Jiangsu includes five cities such as Yancheng, Huai’an, Lianyungang, Suqian, and Xuzhou.) The third is to decompose the TFP growth rate of 13 cities, calculate the capital contribution rate, labor contribution rate, and TFP contribution rate in the economic growth power structure of Southern Jiangsu, Central Jiangsu, and Northern Jiangsu, and analyze the spatial characteristics of urban economic growth power on this basis. Conclusions and policy recommendations are given in the last part. The main innovations of this paper are as follows: Firstly, we put our emphasis on studying the influential elements and constitution of growth rate of TFP in Jiangsu urban economy and the evolution trend of TFP growth rate, while adopting the transcendental logarithmic stochastic frontier model and panel data of 13 cities in Jiangsu. Secondly, we calculate the contribution rate of capital, labor, and TFP to urban economic growth and explore the major influential factor of forming regional gap in terms of urban economic development, so as to provide new thinking in increasing urban TFP growth rate and promoting transformation of driving force for urban economy.
2. Analytical Framework and Methods
2.1. The Estimation and Decomposition of TFP Based on SFA
Based on [1] residual method, TFP is considered to be the residual which equals output growth minus input factors growth. In this method, producer is considered to realize the optimal productive efficiency, and the growth of TFP comes from technical progress, taking no account of technical invalid item and stochastic disturbance. Farrell [19] put forward the idea of frontier analysis efficiency, in which the interference stochastic factors had on output was taken into account, and this idea was of prominently practical significance. Battese and Coelli [20] established a time-varying stochastic frontier production functions model while adopting panel data and a time-varying stochastic frontier production function model (BC95 model) while adding environment variable. In this paper, the TFP of urban economy in transitional period is comparatively objective, which is calculated by adopting the frontier model of transcendental logarithmic production function, in which model of the interference effect on output from stochastic factors is taken into account, the assumption of constant substitution elasticity is relaxed, and the applicability and concrete form could be verified through a series of tests.
2.2. Estimation of Production Function and Decomposition of TFP
At present, models which are applicable to panel data stochastic frontier model are time-varying stochastic frontier production function model by [20] and time-varying stochastic frontier production function which includes model environment variable by [21]. The concrete forms of models by [20] are as follows:
In this model, the functional expression (1) is the stochastic frontier production function, and the functional expression (2) is the time-varying technical inefficiency function. In this model, Yit means economic output of i city in the year t, Xit input vector of i city in the year t, β solve-for parameter, ƞ time-varying parameter of technical efficiency, υit − µit combination of error items (ɛit), υit stochastic disturbance term, which measures the degree of nonefficiency of the system and follows standard normal distribution, µit the error terms of technical loss, which reflects the degree of technical nonefficiency and follows zero cutoff half-normal distribution, and η solve-for parameter, with η > 0, η = 0, η < 0 meaning technical inefficiency decreases progressively, sustains stability, and increases progressively, respectively. Krugman [22] thought that TFP was composed of technical progress, transformation in scale effect, transformation in allocation efficiency, and transformation in technical efficiency. We take the derivation of t from the production function and drop the subscript for brevity:
From the above decomposition formula, we can get
In the above formulas, , , , and represent, respectively, the growth rate of TFP, the change rate of production efficiency, the rate of technological progress, and the improvement of scale efficiency. represents the growth rate of the input j of i province and j is K or L. E presents elasticity of scale, .
2.3. Decomposition of Contribution Rate of Factors to Economic Growth
Formerly, many scholars used the TFP change rate to economic growth rate as the contribution rate to measure economic growth of TFP, but some scholars such as Yu Yongze pointed out that, for China, whose economic growth rate is comparatively high, the abovementioned way could result in major measurement error; thus, in this paper, we adopt the approximate decomposition of the following production function, with f (Xit)t meaning output level when input factor is in t period, and the input of factor i in t period, i = 1, 2, 3, …; t = 1, 2, ….
When the period changes from t to t + 1, economic growth can be approximately decomposed as follows:
Though formula (7) is approximate decomposition, the remaining shares small share in economic growth, and we will no longer conduct a study on it in this paper. We set and as contribution rate to economic growth of Xi and TFP, respectively, when the period is t + 1, and their percentage terms can be expressed as formulas (8) and (9), respectively.
2.4. Variable Selection and Management
We choose output data, input data, and factor cost data of 13 cities in Jiangsu province from the year 2000 to 2020 as the study sample, with all the variables limited to urban area.
2.4.1. Data of Gross Output
Data of gross output are from the public data of Jiangsu Statistical Yearbook. We use urban GDP to measure urban gross economic output and take the year 2000 as the base period to measure urban economic output with urban gross value of production, and then we deflate the annual normal GDP of all the prefecture-level city and get the real output.
2.4.2. Data of Capital Sum (K)
The calculation of physical capital stock in base year of every city is the core for calculation of capital stock. In view that it is very difficult to collect capital stock in base year of every city, in this paper, we take the ratio of GDP in 2000 of every city to the GDP of Jiangsu in 2000 as the weight, distribute the capital stock of Jiangsu in 2000 proportionately to every city, and then calculate the physical capital stock of every city in base period.
After we confirm the physical capital stock of base year, we can calculate annual physical capital stock of every city in Jiangsu with perpetual inventory method at constant price (in the year 2000) with 10.6% yearly depreciation. (Different scholars have different evaluation method for capital stock. In this paper, we adopt the rate of depreciation of 10.96% by [23] to evaluate the capital stock of all cities.)
2.4.3. Data of Labor
We choose arithmetic mean value of “employee numbers” at the end of the year t − 1 and t as labor force input of the city in the year t.
2.4.4. Environment Variable
We choose PGDP, STRU, SCALE, OPN, RD, and ROD as the environment variables affecting production efficiency. We use per capita GDP to measure PGDP, proportion of secondary industry in GDP to measure urban STRU, the total urban population at the end of the year to show SCALE, the proportion of total exports in GDP to show OPN, with US dollar exchanged into RMB at current rate, regional research and development input intensity to show RD, and per capita paved road area to express proxy variable of ROD. Per capita GDP and total exports have been evaluated and deduced.
2.5. Model Test
We use likelihood ratio statistic to test the applicability of stochastic frontier model. For simplicity, we give the test result as follows directly. Table 1 is the test result of model hypothesis, which shows that the concrete form of translog stochastic frontier production function is as follows.
Table 1, the test result of model hypothesis, shows that the concrete form of translog stochastic frontier production function we adopt in this paper is as follows:
3. Calculation and Analysis on Production Efficiency of Different Regions in Jiangsu
3.1. Multicollinearity Test and Estimation of SFA Model
To analyze the influence factors of production efficiency function, we adopt one-step method and two-step method to estimate the model, with model one corresponding to two-step method and model two one-step method. In model one, the multicollinearity of 6 influence factors will be tested. And the inefficiency regression model is as follows:
Collinearity index is VIF = 3.45, and through multicollinearity test, the maximum value of variance inflation factor of influence factor is 7.72, less than 10, the reasonable value, which means that the applicability of the two-step model we use to estimate production function is good. In model two, we use the actual GDP of 13 cities as the output (y) and capital stock of all the past years (k) and labor force (l) as the factor input, add 6 efficiency influence variables, and adopt the software Frontier 4.1 for regression of stochastic frontier model of production function. We get the following result.
From Table 2 measurement results, the value of γ in model one is 0.997, which means that technical inefficiency results in producers deviating from the frontier. In model two, influence factor of technical nonefficiency item is added, and the value of is 0.639. On the whole, the explanatory power of model two increases by 35.9%. The growth rate of urban production efficiency in cities with high production level shows a Mathew’s Effect, which means that cities in South Jiangsu that have advantages in infrastructure, positions, elements, and policies have the abilities to increase production efficiency. There is a negative correlation between the high class way of industry structure and production efficiency. The reason may be that during the process of pursuing high class the industry structure has sublated the negative sides, realizing the promotion of production efficiency, but resulting in decrease in scale effect of original industry. There is positive correlation between city scale and production efficiency, which means that the increase in city population promotes the urbanization level. High level in urbanization means advantages in industrial technology, labor to force, element reallocation, and policies. The efficient reallocation of resources at last promotes production efficiency. High level in opening up could promote the regional production efficiency. The promotion of production efficiency gives enterprises more chances and more confidence in participating in international competition. For export-oriented enterprises with high efficiency, the favorable reallocation of market shares, competitive advantages, and whole promotion of study effect during process of enterprise competition promote the regional production efficiency. In model one and model two, the coefficients of research and development input intensity of different regions are remarkably different, which shows that the correlation between research and development input intensity and production efficiency is not obvious. The constant improvement of city infrastructure has positive externality for urban economic development and production efficiency promotion. This is verified in this paper. It means that the constant improvement of infrastructure such as city roads could promote element efficiency in urban economics, decrease trade cost, and promote concentrated development of urban economics.
3.2. Calculation of TFP Growth Rate
In this paper, we use Frontier 4.1 to estimate SFA model. In equation (4), we get the value of decomposing indices of TE, TP, and SE from the year 2000 to 2020. In Figure 1, we have the TFP growth rate trend in South Jiangsu, Central Jiangsu, and North Jiangsu. From the change trend of the TFP growth rate, the TFP growth rate in the three regions of Jiangsu shows a changing trend of W and shows sharp decrease in 2008 and obvious rebounds in 2013 around. Around the year 2007, the TFP change trends of all cities in Jiangsu basically match each other, which shows that the urban economic efficiency in Jiangsu is affected greatly by macroeconomic situation and national policies.
4. The Decomposition and Spatial Characteristics of TFP Growth Rate of Jiangsu Cities
4.1. Decomposition of TFP in Jiangsu Cities
On the basis of measurement model two, we decompose TFP growth rate index and could get TFP growth rate and change rate of production efficiency, technical development, and scale efficiency improvement through calculation. Figure 2 shows the changing trend of TFP growth rate and its decomposition of cities in Jiangsu. From the time dynamic features of decomposition, the fluctuation range of scale efficiency improvement among 2000 and 2020 is low, except the peaks of wave of scale efficiency improvement appearing in 2003 and 2010; in other years the fluctuation rage of scale efficiency improvement shows slight reduction, which means that the investment expansion could result in decrease in marginal efficiency, that scale efficiency could be improved at certain degree, that the growth rate of production efficiency shows changes between ups and downs between the section of (−0.034, 0.008) and shows a zigzag growth trend, and that the growth rate of technical advance is remarkably higher than other elements, showing that all regions in economic development have strengthened technical import and promoted utilization efficiency. The growth rate of production efficiency and scale efficiency has less influence on TFP growth rate, and the contribution of improvement of scale efficiency to TFP growth rate fails to get the expected result. Through the analysis, we could conclude that technical advance is the major cause of TFP growth in cities of Jiangsu and that the changes of production efficiency growth rate and scale efficiency in low section offset the effect of technical advantage for economic efficiency.
4.2. Structure Features of TFP Growth Rate in Jiangsu Cities
Table 3 shows the accumulative change situation of city TFP growth rate, change rate of production efficiency, technical progress rate, and scale efficiency. From the changing trend of production efficiency, the accumulative changes of urban economic efficiency on different regions are remarkably different, and their common characteristic is that the TFP growth relies mainly on technical advance (TP). The TFP growth rate and decomposition structure index of South Jiangsu lie at the leading level in Jiangsu, which means that the economic development lies mainly on innovation-driven development way. The TFP growth rates in Middle Jiangsu and North Jiangsu lie mainly on technical advance (TP), and their production efficiency (TE) and scale efficiency (SE) show negative growth, decreasing the TFP growth rate in the two regions. One reason why technical advance has not brought about an obvious promotion in production efficiency and scale efficiency may be that technical innovation promotes technical progress, which makes production technology frontier move forward, but could not ensure the efficient promotion of competition efficiency and scale efficiency of the whole industry. Besides, during the process of inviting investment appearing in Middle Jiangsu and North Jiangsu, many input elements are attracted and accumulated there, and the increasing intensity of element input, instead, results in the decrease in scale efficiency.
4.3. Time Characteristics of Driving Force of Jiangsu Urban Economic Development
From equations (8) and (9), we calculate the results in different periods of economic development contributions of different elements in provincial urban economic development as in Tables 4 and 5. In this paper, we will analyze the periodical characteristics of economic development driving force in different regions. The first period is from the year 2000 to 2004, in which the change of TFP growth rate is comparatively stable. In a whole, during this period, the contribution of capital, labor, and TFP to economic development is comparatively steady. The second period is from the year 2005 to 2008, in which the economic development experienced a new round of rapid development.
The influence of financial crisis in 2008 on economics began to show, and there appeared major fluctuation in urban TFP, but the contribution of TFP to economic development was higher by 5.841% than that in the first period. The third period is from the year 2009 to this day, the driving effects of capital input for urban economic development have been strengthened, and at the same time, the contribution of TFP to urban economic development shows sharp decrease. In the circumstances that our nation strongly initiates the policies of “developing new driving force for economic development,” “mass entrepreneurship and innovation,” and “vigorously transforming the pattern of economic growth and achieving high-quality development,” after 2013, the contribution of TFP rebounded back a little.
4.4. Spatial Characteristics of Driving Force of Jiangsu Urban Economic Development
4.4.1. Economic Development Driving Force Features of South Jiangsu
The spatial characteristics of economic growth driving force in South Jiangsu are similar to that in the whole province, which are as follows: firstly, capital input is the main driving force for economic growth, which is below the average level of the whole province; secondly, the contribution level of labor forces is low, which shows a stable condition and grows up slowly in the second period; thirdly, the contribution of TFP to economic growth is lower than that of capital element, but higher than that of labor force. Comparing with the average level of the whole province, the contribution of TFP to economic growth is comparatively high in South Jiangsu but showed a sharp decrease after the year 2009 and bounced back in 2011, which was earlier than in other regions in Jiangsu.
4.4.2. Economic Development Driving Force Features of Middle Jiangsu
The economic growth driving force in Middle Jiangsu shows little difference compared to that of the whole Jiangsu province, which are as follows: firstly, capital input is also the major driving force for economic growth, whose contribution level to economic growth is close to that of the average level of the whole province, becoming the biggest driving force for economic growth in Middle Jiangsu; secondly, the contribution of labor force is comparatively low, a little more than the average level in the whole Jiangsu province in the third period, lower than the average level of the whole province in other periods; thirdly, the changing trend of TFP contribution to economic growth is basically the same as that of the labor force, a little lower than the average level of the whole province in the first two periods, and higher than that of the average level of the whole province in the third period. But it showed a sharp decrease after the financial crisis in 2008 and bounced back a little after 2013.
4.4.3. Economic Development Driving Force Features of North Jiangsu
The changing trend of economic growth driving force in North Jiangsu is shown as follows: firstly, the contribution of capital input is apparently higher than that of the average level of the whole province, being the biggest driving force for economic growth; secondly, the contribution of labor force is comparatively low, which is higher than that of the whole province; thirdly, the contribution of TFP to economic growth is apparently lower than that of the average level of the whole province, showed a sharp decrease after the financial crisis in 2008, and bounced back obviously in 2010 and 2013.
In a whole, the contribution of TFP to economic growth in South Jiangsu is the biggest among the three regions, and then is that in Middle Jiangsu, and the last one is that in North Jiangsu. One reason may lie in that in South and Middle Jiangsu there are more advantages in industrial base, technology, innovation, and talents. In the years from 2009 to 2012 after the financial crisis, due to sharp decrease in TFP in South Jiangsu, the contribution of TFP to economic growth deceased apparently, which was gentle in Middle Jiangsu, and little in North Jiangsu, the reason of which lies in that in South Jiangsu there are lots of export-oriented enterprises, and to lessen the shock from financial crisis, the government adopted economic stimulus policies to pull up the contribution of capital input, and this kind of extensive investment resulted in decrease in economic efficiency, bringing down the contribution of TFP to economic growth.
5. Conclusions and Suggestions
By adopting stochastic frontier transcendental logarithmic production function model, we explore the regional characteristics of economic growth driving force in cities of Jiangsu by calculating the production efficiency and TFP from the year 2000 to 2020 of the 13 cities in Jiangsu and decomposing the changes of TFP from production efficiency growth rate, technical progress rate, and scale efficiency. The study has filled the research gaps in the past, which did not use stochastic frontier model of translog production function to calculate TFP growth rate of a region. This paper also provides a new algorithm for measuring the driving force of regional economic growth such as the driving force of TFP.
5.1. Conclusions
(a)We adopt two-step method to explore the stochastic frontier model parameter and find that the global explanatory power of the model is enhanced by 35.9% after the effect variable of the production efficiency is changed. The urban economic development level, urban size, opening-up degree, and infrastructure construction level can apparently promote the growth of production efficiency. There exists negative correlation with the economic structure and production efficiency. The effect of research and development input intensity on production efficiency is uncertain.(b)There is significant regional heterogeneity in TFP growth rate of Jiangsu. The TFP growth rate for South Jiangsu is higher than that in other regions of Jiangsu province and takes the leading position. The growth trend of TFP in cities of Middle Jiangsu is close to that of the average level of the whole province. The TFP growth rate for North Jiangsu is in the third echelon. Fuglie [12] also confirmed that the growth rate of TFP in China’s developed areas was faster than that in other developing areas. In decomposition of TFP, the contribution to technical progress of TFP growth rate is the biggest, showing a slow rising trend in the process of fierce changes. This view has been further verified in the relative literature. Mariyono [24] support that technical progress is the most important driving force for TFP growth. In 2009, the TFP showed a sharp decrease. The reason is that the decrease in elements results in the significant decline in TFP growth rate.(c)The contribution of capital to the urban economic growth occupies the first position, and the TFP is the second, and labor element is the third. Clearly, economic growth style of Jiangsu urban economy is a classical capital input style, by which the input lies in a leading position in urban economic development. Ahmed [7] reached similar conclusions in studying the productivity growth of manufacturing industry in Malaysia. Especially after the financial crisis in 2008, the government participated in urban infrastructure construction and carried out capital support projects for key industries by means of growing fiscal capacity, which made the contribution of capital to economic growth even more obvious [3, 6]. Besides, the fluctuation features of TFP have the similar trend to economic fluctuation features, and the TFP shows the same changing trend with that of economic development. Benzaquen [16] also got similar findings and believed that the evolution of TFP of the shipbuilding industry coincided with the GDP expansion and decline periods in Peru. From the data, the period in which the contribution to TFP is comparatively large is around the year 2007, when the economic development in China was in high growth period. However, with the constant progress of industrialization and urbanization, the marginal effect of capital is becoming less and less. At the same time, with the disappearing of demographic dividend, the resource allocation efficiency is low, and the conditions on which the traditional urban economy develops have changed, so the transformation of the economic development style which depends mainly on TFP is extremely urgent.5.2. Policy Implication
In the future, to promote the total production efficiency and to transform the economic growth driving force in Jiangsu, we should work on the following points:(a)Because urban economic scale, urban size, opening-up degree, and infrastructure level are conductive to the growth of production efficiency, cities should provide convenient infrastructure, cultivate more human resources, and attract more capital and technology to urban clusters, so as to realize high-quality urban development and drive the continuous improvement in production efficiency.(b)In order to realize the positive and stable growth of total factor productivity, we must give full play to the basic advantages of Jiangsu’s rich technological innovation resources and establish an innovation ecosystem suitable for high-quality development. We should encourage enterprises to carry out innovation and entrepreneurship activities, build a fair trading platform, encourage the industrialization of a large number of innovative achievements, and let innovators enjoy high returns. For the regional heterogeneity distribution of Jiangsu urban regional innovation factors, we should gather them in the urban areas with advantageous factors to control the investment of medium and low-end alternative industrial factors.(c)The transformation of economic growth power mainly relied on TFP needs to improve the level of urban TFP growth rate of Jiangsu. At the same time, it makes up for the lack of TFP growth rate of Northern Jiangsu. Therefore, for the low growth rate of TFP in North Jiangsu, we must vigorously encourage enterprises to carry out independent innovation, build an industrialization platform for innovative achievements, continuously improve the role of technological progress in TFP growth, and realize innovation-driven urban economic growth in Jiangsu. North Jiangsu provides attractive policies in terms of resources, talents, and technology to attract high-quality innovation achievements in South Jiangsu, which can be industrialized in North Jiangsu, improve innovation elements and innovation ability in North Jiangsu, transfer them to urban clusters, and break the weakness for technological progress of North Jiangsu in TFP growth rate.
5.3. Limitations and Suggestions for Future Studies
Research limitations: the research objects are 13 cities in Jiangsu province of China. The research logic is rigorous and scalable, but the resource endowments are different from different provinces, and the research conclusions may be different. Especially in the provinces with serious pollution, how to bring the pollution emissions into the research model and get a more real TFP will be the focus of further research. Due to the lack of research on the dynamic characteristics of economic growth in 13 cities in Jiangsu province, there are few domestic confirmatory literature pieces, so we can only use foreign similar literature to confirm the reliability of the research conclusions.
Suggestions for future studies: when other scholars conduct similar research, we advise them to select areas with strong corroboration for research, compare with the research conclusions of other scholars, and further improve the reliability of research conclusions. When studying the total factor productivity of provinces with serious pollution, such as Hebei Province and Shanxi Province, it is suggested to learn from Ahmed’s scheme (2008), internalize pollution emissions, and incorporate them into the research model, which may get a lower TFP contribution rate. If conditions permit, it is suggested to select DEA-Malmquist algorithm and super logarithm stochastic frontier production function method for confirmatory comparative research to improve the reliability of TFP calculation conclusion.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This study was supported by the Jiangsu Social Science Fund Project Social, Grant no. 19EYD006, and National Social Science Foundation in China, Grant no. 19BJY001.