The main factors for sustainable economic development include: factor accumulation, productivity, and basic factor [26]. Factor accumulation is comprised of physical capital and human capital; productivity mainly refers to total factor productivity, the increase in output results from technological advance except for capital and labor input; basic factor mainly refers to the government’s role in sustainable economic development and the construction of the legal system. Based on this, a model as follows is designed for testing: (1)

In this model, y_i,t is the real GDP of region i in t years, which is calculated by the constant price of 1997; I_i,t is the index for the institutional improvement of region i in t years. This paper focuses on the impact of government governance and legal environment on sustainable economic development. Marketization, the relationship between government and market, development of market intermediaries and the legal environment by [23] are used to survey the vicissitude of institutional environment, government governance and legal system. These index have been widely used in the literature (e.g., [27,28]) to measure China’s regional institutional development, government governance and legal system. Among them, the marketization index captures the following aspects of regional market development: (1) relationship between government and markets, including five sub-index: proportion of economic resources allocated through markets, tax and charge burden on farmers, government’s excessive administrative intervention in enterprises, enterprises’ nontax burden and excessive size of government. (2) development of non-state sectors, including three sub-index: proportion of the non-state sector in gross industrial output, proportion of the non-state sector in national investment in fixed assets and proportion of the non-state sector in number of employed persons. (3) development of product markets, including two sub-index: percentage of products with market-regulated prices and degree of local protectionism. (4) development of factor markets, including four sub-index: development of the financial market, degree of absorption of foreign investment, liquidity of the labor force and transactions of local technology market weighted by number of technicians. (5) development of market intermediaries and the legal environment, including four sub-index: development of market intermediaries, protection of producers’ lawful rights and interests, protection of intellectual property rights, protection of consumers’ interests. According to the theme of this paper, we focus on three indexes and their sub-index: marketization, relationship between government and markets and development of market intermediaries and the legal environment. control is the control variable, and according to the existing literatures, it mainly includes natural logarithm of physical capital stock , ln k_i,t natural logarithm of human capital stock ln h_i,t and productivity TFP_i,t. Among which, K_i,t is the physical capital stock of region i in t years calculated from the formula K_i,t = K_i,t–1 (1–0.096) + I_i,t according to the method of Zhang et al. [29]; K_i,t–1 is the physical capital stock of region i in t–1years, initial capital stock is the physical capital stock of 1999, which is calculated by constant price of 1997. I_i,t is investment volume of fixed capital, which is calculated by the constant price of 1997, and the rate of depreciation is 9.6%. h_i,t is the human capital stock of region i in t years, which is calculated by labor force’s average years of education multiplied by the number of employees, which leads to the formula that labor force’s average years of education equals illiteracy × 0 plus primary school × 6 plus junior high school × 9 plus senior high school × 12 plus junior college × 15 plus college × 16. TFP_i,t is the total factor productivity of region i in t years, which is calculated according to the method of Zhang and Shi [30], namely, TFP_i,t = Y_i,t /( × ).

In this paper, the number of employees and labor force’s average years of education are derived from the China Labor Statistical Yearbook; institutional variables from Annual Report 2011: Marketization Index for China’s Provinces; and other data from Annual Data of China Regional Economy Database in China Stock Market and Accounting Research Database(CSMAR), this database is prepared by the China Accounting and Finance Research Center of the Hong Kong Polytechnic University and the Shenzhen GTA Information Technology Company limited. The period that has been surveyed is 1999–2009. Although the indexes designed by Fan et al. [23] start from 1997, this paper, however, only analyzes the 1999–2009 data to find out the impact of government governance and legal environment on sustainable economic development because the data of 1997 and 1998 are incomplete in many provinces. Meanwhile, because of the obvious regional disparity of Chinese economic development, the paper divides the 31 provinces, municipality, autonomous region in China into Eastern region, Middle region and Western region according to their economic development level. Among them, Eastern region includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan, Middle region includes Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hunan and Hubei, Western region includes Inner Mongolia, Guangxi, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang, Sichuan and Chongqing (exclude Tibet because of data limit).

In light of the endogenity of institution, the paper selects instrumental variable method to make a regression analysis on Formula 1. The enrollment ratio of secondary school by [31] is selected as an instrumental variable. The calculating data of this, the number of secondary school’s enrollment and the number of graduates from primary school, are derived from the Statistical Yearbook for China’s Regional Economy.

Table 1 shows the descriptive statistics results of all variables. It shows that there are differences among total factor productivity, marketization, relationship between government and markets, and legal environment in China’s 30 provinces, municipalities and autonomous regions.

As for the evaluation and design of panel data model with instrumental variables, this paper proceeds as follows: first, F Test is used to select fixed effect model with instrumental variables or regression model with instrumental variables; if the fixed effect model with instrumental variables is selected, then a fixed effect model with instrumental variables or a random effect model with instrumental variables shall be selected according to the results of Hausman Test; if the selection is still the fixed effect model with instrumental variables, the test is then finished; if the random effect model with instrumental variables is selected, LM Test on the selection between random effect model with instrumental variables and regression model with instrumental variables should be done as a complementary test.

It is shown in Table 2 and Table 3 that the results of panel regression on government governance, legal environment and economic development by inter-provincial panel data. It is indicated in Table 2 that physical capital stock, human capital stock and total factor productivity have significant positive effects on sustainable economic development, and they have passed a 1% significance test. After the institution variable is added, which means the indexes of marketization, relationship between government and markets, development of market intermediaries and the legal environment are added into the model, physical capital stock, human capital stock and total factor productivity still have significant positive impact on sustainable economic development. In other words, physical capital stock, human capital stock and total factor productivity are the main factors in China’s sustainable economic development. Meanwhile, we find that each index of the institutional variables shows a significant positive impact on sustainable economic development, namely, it plays a significant role in promoting sustainable economic development, which is consistent with the conclusions of existing research literatures [32,33,34].

Then, which institutional factors have effects on sustainable economic development? It is shown in Table 3 that the results of panel regression on each index of relationship between government and markets, development of market intermediaries and the legal environment and sustainable economic development of inter-provincial panel data. Physical capital stock is still significant in all models, while human capital stock and total factor productivity is not significant in only one model; meanwhile, the results of the relationship between government and market indexes are surprising: the proportion of economic resources allocated through market shows a significantly negative impact on sustainable economic development, namely, the more economic resources are allocated through market, the less economic development would be seen, which totally run counter to the theory of economic development. Considering the symbol of the index of reducing government’s excessive administrative intervention in enterprises, it can be seen that the market economy has been implemented by the Chinese government since 1994, but China’s enterprises are still difficult to adapt to the operating environment of market economy because they are used to receiving economic resources allocated through the government. The index of reducing the excessive size of government shows a negative impact on sustainable economic development, but not a significant one. It indicates that the government scale cannot be reduced blindly, and the key is the scope of the government’s jurisdiction; the efficiency of the civil servants rather than the numbers of them. The two indexes of reducing the tax and charge burden on farmers and reducing the enterprises’ nontax burden show a significantly positive impact on sustainable economic development, which confirms the promotion effect that the more government’s reducing the tax and charge burden on farmers and enterprises’ nontax burden, the better sustainable regional economy would develop. Models 6–9 show that the impact of legal environment’s indexes on sustainable economic development passes a 1% significance test. Among which, the development of market intermediaries, the protection of the producers’ lawful rights and interests and the protection of consumers’ interests, and their sign is positive, which means the better these factors are, the faster the sustainable economic development is. However, the relationship between protection of intellectual property rights and sustainable economic development is in a negative correlation, which means the better protection of intellectual property rights is, the slower the sustainable economic development is. It may result from a lower level of economic development in China. An empirical study of Thompson and Rushing (1996, 1999) find that in developed countries, patent protection can effectively promote the total factor productivity; while in developing countries, the implementation of minimum standards for the patent protection would not promote economic development, unless the economic development of the country has reached a certain level [35,36].

For a long time in history, Midwestern China, especially Western China is fairly secluded, where the dominant institution shows characteristics of Chinese traditional feudal natural economy: physiocracy, value just above material gains, individual’s subjecting to the government and authority, ignoring the rule of law. Therefore, compared with Eastern China, Midwestern China has a less developed commodity economy and lacks the awareness of commerce. Meanwhile, China’s reform is a gradual process of institutional transformation. According to the reform plan of the central government, the reform shall be implemented in Eastern Coastal Regions first as a pilot reform, and then is introduced to the Midwestern China. These regional differences in the order of the formal and informal institutional reforms together determine the differences in China’s regional sustainable economic development. Do the above-mentioned conclusions still have strong robustness after taking these factors into account? To this end, this paper respectively surveys the impact of each institutional index in the eastern and western regions on regional sustainable economic development.

Table 4 shows the regression results of government governance, legal environment and sustainable economic development by the Eastern and Western panel data. It is indicated in Table 4 that those indexes of marketization, relationship between government and markets, development of market intermediaries and the legal environment in Eastern China are consistent with the regression results in inter-provincial panel data, which shows a significant positive correlation. It confirms the view that a better legal system leads to better economic performance. While those indexes of marketization, relationship between government and markets and legal environment are still consistent with the regression results in Eastern China, they are not significant in Western China. The significant impact of marketization, relationship between government and markets and legal environment on sustainable economic development has not yet been shown completely. Therefore, further improvement of the government governance and legal system in Western China plays an important and critical role in sustainable economic development. The conclusion is consistent with the results of [37]’s cross-country data, whereby they found that government governance does not bring significant economic growth when countries are in a lower income stage, but in higher income stages, there is a strong correlation between government governance and economic output.

Table 5 shows that, in Eastern China, the indexes of relationship between government and markets and sustainable economic development are still the same with the sign of variables in inter-provincial panel data, but the levels of significance are different. The index of proportion of economic resources allocated through market and reducing enterprises’ nontax burden have passed the 10% significance test; the index of reducing tax and charge burden on farmers has passed the 1% significance test. The index of reducing government’s excessive administrative intervention in enterprises and reducing excessive size of government has not passed 10% the significance test, which indicates the effect of government governance still needs to be tested, and the significant effect is still not shown completely. In Western China, the impact of the indexes of relationship between government and markets on regional sustainable economic development show a great difference with that in Eastern China. All the indexes show a positive correlation but do not pass the 10% significance test, which indicates that the impact of the indexes of relationship between government and markets on regional sustainable economic development is limited in Western China.

Table 6 shows the impact of development of market intermediaries and legal environment’s indexes on sustainable economic development in Eastern China. These results are the same as the regression results of China’s inter-provincial panel data. However, while the protection of intellectual property rights shows a significantly negative impact on regional sustainable economic development, other indexes show a significantly positive impact. In Western China, while the index sign of the protection of producers’ lawful rights and interests is negative, the sign of development of market intermediaries, protection of intellectual property rights and protection of consumers’ interests are positive but not significant. This further illustrates that the level of economic sustainable development is determined by the different legal environments in Eastern and Western China. These conclusions confirm [38]’s results that the development of law system (a formal contract subject) will gradually become more and more important as economic development reaches a higher stage.