Chinese / English
Home >> Great Minds China Forum >>  2nd Great Minds China Forum >>  Investment, Mismatch and China's Economic Growth: Challenges and OpportunitiesInvestment, Mismatch and China's Economic Growth: Challenges and Opportunities
Bai Chong’en:The New Dual Structure of Chinese Economy

Bai Chong’en


I am obviously not a neutral commentator as I conducted many research with Song Zheng. My topic is The New Dual Structure of Chinese Economy.


In this morning, the two award winning professors introduced their innovative researches on the institutional issues in China’s economic development which inspire us a lot. They depicted the situation in China and also made contributions to the economic theory.


In terms the mode of economic growth, Professor Qian and Professor Xu talked about the comparison between M-form and U-form this morning. They said M-form was beneficial to the holding of GDP Championships. This is a very important opinion that explains many issues and phenomena in China. There is room for us to explore. If there is a GDP Championship, local governments might do many things like improving the market environment to enhance the growth potential of the local GDP under pressure of competition. If we believe in the findings of institutional economics and institutional environment will positively influence the economic growth, local governments should try hard to improve the local market environment. For market environment, a significant point is that if there are GDP Championships, all enterprises should be treated equally and enterprises could have preferential offerings. However, if local governments are owners of part of the enterprises, they will help enterprises to protect property right. Overall, the goal is to seek for local GDP growth. Improving local business environment is a crucial method. Therefore, I think the local government is able to offer such help like getting you pass the environmental assessment when you failed or helping you get the land you want when you faced with great challenges. Local governments are capable of offering help. This might due to the GDP Championships or hoping to be praised on political performance as having a model enterprise can be good when leaders come for inspection. This might also due to personal benefits or political, reputation, economic or spiritual considerations as choosing some enterprises could provide him with the greatest benefit on politics, spirit or economy. They will select some enterprises to help them tackle institutional obstacles. We can define what enterprises do as preferences, improving overall business environment as generalized preferences and solving enterprises’ issues case by case as preferences. We call subjects that receive preferential offerings as government linked enterprises. Such name is used in Singapore. Why don’t we call them state-owned enterprises (SOEs)? That’s because many or even most of the government linked enterprises are not SOEs. There are no local governments that hold the share. Completely private enterprises can also be government linked enterprises. For constructing the model, we can divide enterprises into two types, one is government linked enterprises and the other is non-linked enterprises. These two types of enterprises form the dual structure. When we discuss dual structure, we say that rural and urban areas are dual structures which is still crucial thing. However, if we want to have insight into Chinese economy, we have to thoroughly understand the dual structure of government linked enterprises and non-linked enterprises.


What I want to say is that even though these two theories about the mode of Chinese economic growth, they are not mutually exclusive. They are different but they have close correlation or even relationship of inheritance.


Just as we said, if local governments choose some enterprises and help them get across some difficulties, what are the consequences? Local governments control many resources and they can help enterprises from various perspectives which could generate different results. We divided the way of help into two types. One is to help enterprises to cut the red tapes of imperfect systems. Local government can solve the issue of registration by a simple phone call. For example, if there are several contradicted regulations and enterprises can do nothing if they have to obey them, then we need a local government official to guarantee that it is no problem to disobey some of the regulations as long as not going too far. If the help offered by local government is mainly in this form, then it could be beneficial to the economic growth. The overall institutional environment is not good and all the enterprises are faced with different obstacles. If local government cannot eliminate all the obstacles, it still helps even if they can only remove some and the efficiency can still be improved. This is one type of help.   


The other is to directly provide resources to enterprises. Local governments control different resources. They are in control of land resources. They can put the lands in a financing platform where lands would be used as mortgage. This is a significant phenomenon. By borrowing money, local governments can help whatever enterprises as they want. Injecting resources into enterprises directly is the second type of help. The resources can be capital, land, etc. Due to limited resources, when local government put resources into one type of enterprises, the other type of enterprises will have difficulty in acquiring resources. When limited resources are used by one party, the cost of acquiring them would raise. The second type of help will decrease the efficiency as it generates distortions. If there were two enterprises, one is government linked enterprise and the other is not, why did you offer more capital to the government linked one? This should not happen in terms of efficiency. But it did happen. Therefore, the results of preferential mode depend on what kind of help local governments offer to those subjects.


Song Zheng mentioned many contents, especially the different indexes of effectiveness which had been improving before 2008. However, these indexes were decreasing after 2008. What are the reasons for this huge change? Did it have something to do with the mode of preferential offerings? The answer is yes.


How? Before we revised the Budget Law in 2014, local governments were not allowed to borrow money in a long period of time. However, they could get across this limitation like building a financing platform where all state-owned enterprises of local governments borrow money. Before 2008, there were not many financing platforms and they were in small scale. In addition, the regulations were quite strict. At that time, we knew that local governments could do many things which not necessarily beneficial to the economy if they have too many resources. Local governments were eager to make investment. Things were different at the end of 2008. We encountered the shock of international financial crisis. At that time, we prioritized the economic growth. Local government did not eager to invest any more. We worried that local governments no longer want to invest or invest too little. Therefore, the policy and regulation on financing were loosened. From 2008 till now, especially between 2009 and 2010, the financing platforms sprung up like mushrooms after the rain. Financing platforms made an increasing amount of loans. By directly selling resources like lands through financing platforms, local governments acquired many financial resources. In the past, local governments helped subjects who received preferential offerings by helping them to pass the environmental assessment or many regulations. However, after 2008, they can directly or indirectly offer capital support. Such great change has resulted in the change of the nature of help in the mode of preferential offerings. The first type of help could increase efficiency while the second type could increase distortions.


Lets assume that we divide economy into government linked enterprises and non-linked enterprises these two parts. When you offer help to linked enterprises, they would occupy resources that non-linked ones should get. This is the crowding out effect. Such effect will decrease efficiency. This is the simplest model that we could think of.


Some say that the efficiency of non-linked enterprises could be increased if the investment we make in lined enterprises is with positive externality. If so, helping linked enterprises is beneficial. Some believe that industrial policy can increase efficiency. Sometimes industrial policy does work. However, for most of the time, it fails. If we look at the Japans research, we find that Japanese did comprehensive econometrics research in this field. The profits generated by the successful industrial policy that implemented by Japanese government were much lower than the cost brought by unsuccessful industrial policy. The reason is that when one industrial policy failed, the government would keep supporting it to prove it is correct. That is soft budget constraint. Therefore, the unsuccessful industrial policy will continue for a long time and causes great cost.


We combined external factors with this research model. If we have a policy that opens the economy rather than closing it, other enterprises’ capital cost would increase when we offer preferences to enterprises. Therefore, the capital that should have been invested overseas would be invested domestically. This could increase the capital cost of enterprises, lower their competitiveness in the global market and result in the decrease of trade surplus. These all coincide with our observations.


Together with two of my colleagues, I researched how wages of different workers and workers with different education level were influenced. The basic assumption of this model is the enterprises supported by government are primary labor intensive industries. By doing empirical research, we found that when the government supported enterprises’ demand for primary labor increased, like the employment rate of construction industry reaching a higher proportion, the increase in primary labor proportion is faster than that of skilled labor. We made basic assumption that government levies on consumers to subsidy the interest expenditure of government linked enterprises. Under this mode, the average rate of capital return will decrease; interest rate will increase; skill premium will increase; the ratio of wages of employees with higher and lower education will decrease; consumption will be negatively influenced; output will increase in the short term but not consistent in the long term. These coincide with our observation. Before 2000, the return on investment (ROI) was quite stable but it sharply decreased after 2008. The skill premium of wages was increasing before 2008 then decreased afterwards. The results that predicted by the model coincides with the actual data in many aspects, like the proportion in different department, investment rate and skill premium.


Last but not least, what does this understanding mean to our policy?


Firstly, the institutional changes. For example, the anti-corruption campaign dampens the local governments enthusiasm in providing preferential offerings. How can we motivate local governments to keep supporting enterprises under such circumstances? Can we motivate local governments to lower enterprises’ cost? Chenggang said we need some fundamental changes in the incentives for local governments in order to have some incentive policies of realizing generalized preferences. I quickly mentioned about industrial policy. State-owned enterprise is the representation of industrial policy. What we actually should do is to reduce government’s capacity of providing resources to linked enterprises. Just now, we said the main differences before and after 2008 was because local governments obtained a huge amount of financial resources after 2008. When they helped enterprises, they offered resources which could reduce the efficiency. How should we control this ability? The independence of finance is very important. We hope banks would not completely cooperate with local governments on such things. We also constantly emphasized on enhancing budget management. In terms of the debt management of government, after we revised the Budget Law in 2014 and allowed them to issue bonds, we published a series of documents of managing debt of local governments. Now, we need to manage government assets. When local governments own many assets, even if you can constraint their issuing of bonds, they can still borrow money from financial institutions using their assets as the collateral. If we don’t manage local governments’ assets, we won’t be able to solve the issue of soft budget constraint.


Secondly, how is the ROI of financing platform? It is very low. Here you can see the median and the average value of the ROI of local financing platform. Until 2015, the ROI of enterprises that issued bonds were lower than 2%. This is because the nominal rate and the inflation rate is included. In view of the median, it is 0.5%, less than 1%, which is lower when comparing with the ROI of the whole society.


Thirdly, the environment is changing. What worked well in the past may not apply to the future. We might not be short of infrastructures like we did before. In the past, the low cost of low skilled labor was our advantage. However, the cost of low skilled labor surpasses that of Southeast Asian and South Asian countries. We can no longer compete on this. However, our comparative advantage will gradually change into having skilled intellectual labor. Every year, we have many newly graduated engineers. Though the cost of these engineers can’t be compared with that of Southeast Asian and South Asian countries, it is much lower comparing with that of the America and Europe. As the comparative advantage is changing, can old system fit into the new comparative advantage? We still have long road to explore.


Thank you!





◆please indicate the source if authorized: National Economics Foundation

◆photo:National Economics Foundation