why infrastructure important in China

Journal of Comparative Economics 38 (2010) 34–51

Contents lists available at ScienceDirect

Journal of Comparative Economics

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j c e

Infrastructure development in China: The cases of electricity, highways, and railways

Chong-En Bai a,b,*, Yingyi Qian a,c

a School of Economics and Management, Tsinghua University, Beijing 100084, China b National Institute for Fiscal Studies, Tsinghua University, Beijing 100084, China c University of California, Berkeley

a r t i c l e i n f o

Article history: Received 21 October 2009 Available online 27 October 2009

JEL classification: H44 L9 O14 R42 R48

Key words: Infrastructure Electricity Highway Railway China

0147-5967/$ - see front matter � 2010 Published b doi:10.1016/j.jce.2009.10.003

* Corresponding author. Address: School of Econo E-mail addresses: [email protected] (

1 One commonly used approach is to estimate th infrastructure (for example, Aschauer, 1989; Munnel (for example, Hulten and Schwab, 1991; Tatom, 1991 Morrison and Schwartz (1996) and Lynde and Richmo Li (2005) adopts a third approach and uses the chang return to infrastructure investment. He finds significa and Fan and Zhang (2004) and they both find positi infrastructure increases property value (Haughwout, 2 Ying, 1988).

a b s t r a c t

Bai, Chong-En, and Qian, Yingyi—Infrastructure development in China: The cases of elec- tricity, highways, and railways

This paper considers the development of the electricity, highway, and railway sectors in China, with special emphasis on investment incentives. Statistical summary of the devel- opment of these sectors is offered, followed by a detailed description of the institutional background, including investment and pricing mechanisms. We also analyze investment incentives based on the institutional background and present our estimates of the rates of return to investment in these sectors. It is observed that some of the current practices may serve as useful transitional arrangements even though they are not desirable in the long run. Journal of Comparative Economics 38 (1) (2010) 34–51. School of Economics and Management, Tsinghua University, Beijing 100084, China; National Institute for Fiscal Studies, Tsinghua University, Beijing 100084, China; University of California, Berkeley.

� 2010 Published by Elsevier Inc. on behalf of Association for Comparative Economic Studies.

1. Introduction

There is a large literature studying the importance of infrastructure to economic development.1 However, there is not much systematic research on how infrastructure is developed. Many issues are worth consideration. One of these issues is investment incentives. Infrastructure may yield significant social returns. However, this does not guarantee that investors of infrastructure projects can get sufficient private return. How can one provide incentives for private investment? If there is not sufficient private incentive to invest in infrastructure, are there enough incentives for politicians and bureaucrats to over- come many difficulties to make public investment in infrastructure? To address these issues, one has to understand the insti-

y Elsevier Inc. on behalf of Association for Comparative Economic Studies.

mics and Management, Tsinghua University, Beijing 100084, China. Fax: +86 10 62785562. C.-E. Bai), [email protected] (Y. Qian). e production function with infrastructure as an input. Some find significant returns to investment in l, 1990; Rubin, 1991; Holtz-Eakin, 1994), while others do not find significant returns to such investment ; Munnell, 1992; Kavanagh, 1997). Another approach is to estimate the cost function. Using this approach, nd (1992) both find evidence supporting a significant role of public infrastructure in productivity growth.

es in interregional price gaps before and after the construction of a railway in China to estimate the social nt return to the investment. Other studies on the role of infrastructure in China include Demurger (2001) ve relationship between infrastructure and economic growth. There are also findings that suggest that 002), reduces inventory costs (Shirley and Winston, 2004), and cuts the cost of trucking firms (Keeler and

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 35

tutions surrounding infrastructure. Who is allowed to invest in infrastructure? What are the restrictions faced by the investors? In the case infrastructure is built with private investment, how is the price for the use of infrastructure is regulated? In the case of public investment, what is the environment faced by the politicians and bureaucrats? Without a clear understanding of these institutions, it is impossible to understand the development of infrastructure.

Before we can get the full picture, it is useful to start with a few cases. This paper attempts to contribute to this first step by investigating the development of the electricity, highway, and railway sectors in China. We will not consider urban infra- structure because relevant information is more difficult to find and there are also too many aspects of urban infrastructure for us to cover in this paper. We will not consider the communication sector because its development seems less problematic than other sectors of infrastructures and certain success has been achieved in the sector in many developing countries. Among the modes of transportation, we will not consider waterways, ports, civil aviation, or pipelines. Instead, we focus on the sectors of electricity, highways and railways because they exhibit unique Chinese characteristics and they also offer interesting contrast among them.

For each sector we consider, we first offer a statistical summary of its development in China. We present the capacity and the output of the sector. We not only offer a snapshot of the current situation, we also provide relevant historical statistics.

The paper also provides a detailed institutional background for the development of each sector. We particularly empha- size institutions related to investment incentives. These include investment policies and the pricing of the services. Again, the history of the institutions is presented, and the current situation is described in greater details.

Based on the institutional background, we analyze investment incentives and existing problems in each sector. Some practices are part of the solution to providing investment incentives and at the same time sources of economic costs in terms of distortion of investment decision and high prices paid by consumers, and social and political costs in terms of corruption and income inequality. We discuss both the benefits and costs of these practices.

Finally, we present of estimates of the rates of return to investment for each sector. These estimates provide quantitative indication about incentives for investment outside of government budget.

This paper does not consider a few aspects related to investment in these sectors. One is the cost of acquiring the land needed for the development. Another is the cost of complying with environmental regulations and other regulations. We focus more on the benefits of investment to the investors rather than the costs. The cost issues await future research.

2. The electricity sector

2.1. Statistical summary

The development of the electricity sector has basically kept in pace with the growth of real GDP (Fig. 1 and Table 1). From 1978 to 2007, the annual growth rates of installed capacity and the amount of electricity generated were 9.1% and 9.2% respectively. In the same period, the annual growth rate of real GDP was 9.8%. In 2007, the total amount of electricity gen- erated was 3256 billion kW-h, representing a 14.44% increase over the previous year. Around 83% of the output was from thermal power, 15% hydro power, and 2% nuclear power. The total installed capacity was 713.3 GW.

On the demand side, household consumption accounts for 11% of the total. Industry consumes 76.6% of the total, services 9.8%, and agriculture 2.6%, with industrial consumption, especially that by the heavy industry, growing the fastest.

Given the price structure, the demand and supply of electricity is basically in balance. According to the annual report of the 2007 State Electricity Regulatory Commission, the maximum shortfall of capacity during peak load in 2007 was only 10 GW, a mere 1.4% of total installed capacity.

(Logarithm with 1978 Values Nornalized to Zero)

0

0.5

1

1.5

2

2.5

3

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

GDP Indices

Installed Capacity

Electricity Generated

Fig. 1. GDP, installed capacity of electricity generation, and electricity generated. Sources: 1952–1993 data are from China Electricity Information Net (http://www.sp.com.cn/zgdl/dltj/default.htm); 1994–2006 data are from China Electricity Yearbook (CCOCEY, various years); 2007 data are from National Statistical Summary of Electricity Production and Investment.

Table 1 Installed capacity and electricity generated (1952–2007).

Year Installed capacity (unit = 10 MW) Electricity generated (unit = 100 million kW h)

Total Hydro Thermal Nuclear Others Total Hydro Thermal Nuclear Others

1952 197 19 178 73 13 60 1957 464 102 362 193 48 145 1962 1304 238 1066 458 90 368 1965 1508 302 1206 676 104 572 1970 2377 624 1753 1159 205 954 1975 4341 1343 2998 1958 476 1482 1976 4715 1466 3249 2031 456 1575 1977 5145 1576 3569 2234 476 1758 1978 5712 1728 3984 2565 446 2119 1979 6302 1911 4391 2819 501 2318 1980 6587 2032 4555 3006 582 2424 1981 6913 2193 4720 3093 656 2437 1982 7236 2296 4940 3277 744 2533 1983 7644 2416 5228 3515 864 2651 1984 8012 2560 5452 3770 868 2902 1985 8705 2641 6064 4107 924 3183 1986 9382 2754 6628 4496 945 3551 1987 10,290 3019 7271 4973 1002 3971 1988 11,550 3270 8280 5451 1092 4359 1989 12,664 3458 9206 5847 1185 4662 1990 13,789 3605 10,184 6213 1263 4950 1991 15,147 3788 11,359 6775 1248 5527 1992 16,653 4068 12,585 7542 1315 6227 1993 18,291 4459 13,832 8364 1507 6857 1994 19,990 4906 14,874 210 9279 1668 7470 140 1995 21,722 5218 16,294 210 10,070 1868 8073 128 1996 23,654 5558 17,886 210 10,797 1869 8778 143 6 1997 25,424 5973 19,228 210 13 11,342 1946 9249 144 3 1998 27,729 6507 20,988 210 24 11,577 2043 9388 141 5 1999 29,877 7297 22,343 210 26 12,331 2129 10,047 148 6 2000 31,932 7935 23,754 210 33 13,685 2431 11,079 167 7 2001 33,849 8301 25,301 210 37 14,839 2611 12,045 175 8 2002 35,657 8607 26,555 447 48 16,542 2746 13,522 265 9 2003 39,141 9490 28,977 619 55 19,052 2813 15,790 439 11 2004 44,238 10,524 32,948 684 82 21,944 3310 18,104 505 25 2005 51,667 11,739 39,138 685 106 24,655 3952 20,180 523 2006 64,134 14,857 48,405 685 187 28,310 4167 23,573 543 27 2007 71,256 14,526 55,442 885 403 32,529 4867 26,980 626 56

Sources: the same as Fig. 1.

36 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

2.2. The history of development

The development of the electricity industry can be divided into three periods, with the market forces playing progres- sively more important roles.

The first is the pre-reform period before 1978. At that time, the degree of government control was even higher in the elec- tricity sector than in most of the other sectors. The sector was characterized by two features. One is that there was no sep- aration of government functions, such as formulating and enforcing regulations on entry, pricing, and system security, from business activities, such as electricity generation, transmission, and distribution. Both functions were played by one govern- ment agency, the Ministry of Electricity Industry and local bureaus under the ministry, although the name of the agency changed several times. The second is that all activities were under rigid state plan.

Starting from a very low base, the electricity sector experienced a period of rapid growth. At the end of 1949, the total installed capacity in the whole country was only 1850 MW, generating 4.3 billion kW-h of electricity annually, with annual per capita output of less than 8 kW-h. By the end of 1978, the total installed capacity had increased to 57,120 MW, with annual total output of 256.6 billion kW-h and per capita output of 266 kW-h. The annual growth rates of total installed capacity and electricity generation were 12.6% and 15.1% respectively. Despite of the high growth rates, electricity shortage was always a severe problem. Misallocation as a result of rigid planning was partly responsible for the shortage.

The second period is between 1978 and 2002. In this period, state plan was loosened and government functions and business activities began to be separated in this sector. However, competition was limited and the bureaus of electricity industry and later the electric power corporations played an overwhelmingly dominant role in the sector. There were three milestones in the reform process. The first was a 1985 electricity policy document approved by the state council (SCC, 1985). This document allowed investment in electricity generation from outside of the state budget, including

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 37

foreign investment. At the same time, it also allowed the electricity generated by plants built with non-state-budgetary funds to enter the market. As a result, two tracks with different prices emerged in the electricity sector, one within the state plan and the other outside. The second milestone was the 1987 state council policy of separating government func- tions from business activities in the electricity sector. In each province, an electric power corporation was set up parallel to the government bureau of electricity industry. The 1985 policy of encouraging diverse sources of investment in the electricity sector was reaffirmed. Because the provincial enterprises were rather independent of one another, the market was segmented. The third milestone was establishment of the National Electric Power Corporation in 1997 and the sub- sequent abolishment of the Ministry of Electricity Industry in 1998. After the abolishment of the Ministry of Electricity Industry, the government functions in this sector became part of the portfolio of the ministerial level State Economic and Trade Commission.

The growth rates of the sector varied during this long period. In the period of 1979–1984, the annual growth rates of total installed capacity and electricity generation were 4.9% and 6.0% respectively. In the period of 1984–1997, however, the cor- responding rates were 9.2 and 8.8 respectively. In 1997, the sector experienced surplus of supply, and the government then decided not to approve any new investment in electricity generation for the next 3 years. This turned out to be short sighted. By 2003, 22 provinces experienced severe shortage of electricity supply and electricity had to be rationed again.

The third period started from 2002. The state council issued the System Reform Plan for the Electricity Sector in February, 2002 (SCC, 2002), which shapes the current practices.

2.3. Current practices

2.3.1. Disaggregation The 2002 plan consists of three main components. The first is to separate electricity generation from its transmission and

distribution as well as other businesses. The National Electric Power Corporation was broken up into two transmission and distribution companies separated along regional lines, five electricity generation companies, and four other companies offer- ing engineering and construction services for the electricity sector.

With the implementation of the plan, the number of players has increased significant, but the state still maintains a dom- inant position in the market. At the end of 2007, there are more than 4000 plants with installed capacity above 6 MW in the electricity generation sector. Central government controlled enterprises, including the five companies that spun off from Na- tional Electric Power Corporation 5 years ago, had a 53.95% share of the installed capacity, up 4.43% age points from the year earlier. Local government controlled enterprises had a share of 41%, down 4 percentage points, and private and foreign in- vested enterprises had a share of 6.05%, down 0.16 percentage points.

The transmission and distribution sector is mainly divided between State Grid Corporation of China and China Southern Power Grid, the two companies that spun off from the National Electric Power Corporation in 2002 (SERC, 2008). Each power grid company is a monopoly transmitter, distributor, and retailer of electricity in its region. Generators sell power to the power grid company, and their access to the grid is mostly at the mercy of the latter, even though in some cases access is supposed to be determined by a bidding process.

2.3.2. Pricing policy The second component of the plan is to develop a new mechanism for price formation. The idea is to let the market play a

more important role in price formation in the sector in the long run. The price that consumers pay consists of four parts (NDRC, 2005a, 2005b, 2005c). One is the cost for the distributor to

acquire electricity from the generating plant, the second is the losses in transmission and distribution, the third is the price for transmission and distribution, and the fourth is government surcharge. The price for transmission and distribution is determined by the government based on the cost of transmission and distribution. The cost for the distributor to acquire electricity includes the price paid to the generating plant and taxes paid to the government. The price paid to the generating plants is a two-part tariff. One part is the capacity price determined by the government that is independent of the actual amount of electricity that is purchased, and the second part is a unit price that is either determined by a bidding process or by the government based on industry average cost, depending on whether the region has set up a functioning electricity trading system.

Different consumers pay different prices. Households pay the lowest price, followed by agricultural users. Non-agricul- tural business users pay the highest price.

2.3.3. Regulation authorities The third component of the plan is to reform government functions in the sector. The State Electricity Regulatory Com-

mission (SERC) was formed to regulate the industry but the power to determine the prices in the sector rests with the Na- tional Development and Reform Commission (NDRC). SERC’s main charges are to maintain competition, protect the interests of various stakeholders, enforce price regulation set out by NDRC, and guarantee the security of the electricity system (SERC, 2008). National Development and Reform Commission and its local bureaus are responsible for determining the prices when they are not determined by the market.

Table 2 Investment in fixed assets in urban area by jurisdiction of management and registration status: production and supply of electric power and heat power (unit: billion yuan).

Year Total By jurisdiction By registration status By registration status

Central Local Domestic HK, Macao, Taiwan Foreign State Collective Private

2004 485.4 169.9 315.5 446.3 23.8 15.4 392.4 6.3 8.7 2005 650.3 219.4 430.9 601.6 31.3 17.4 507.7 6 12 2006 727.4 266.2 461.2 682.2 27.4 17.8 571.7 42.4 80.3

Sources: China Statistical Yearbook 2005–2007 (NBS, various years).

38 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

2.4. Investment incentives

2.4.1. Sources of investment Substantial amount of resources are invested in the electricity sector each year. In 2007, 304.2 billion yuan was invested

in electricity generation and 245.1 billion yuan in transmission. Investment in the sector is dominated by state-controlled enterprises. This can be seen from the distribution of installed capacity among various types of enterprises that we discussed earlier. Table 2 also illustrates this point, among others. The National Bureau of Statistics does not provide investment data for the electricity sector alone, but it provides investment data for electricity together with heat power, which should be indicative of the electricity sector. From the table, we can see that most of the investment was made by state-controlled enterprises; most of the funds came from domestic sources; local governments invested more than the central government.

As Table 3 shows, of the total invested amount, domestic loan is the most important component. Self-raised funds follow closely behind. Self-raised funds include extra-budgetary funds from central government ministries and local governments, as well as the self-raised funds of enterprises and institutions. State budget is not an important component of the total investment in the electricity sector. Note that Tables 2 and 3 show different amount of total investment. This is because the two sets of statistics come from different sources and the differences between them are statistical discrepancies.

2.4.2. Incentives and problems A natural question to ask is: why are there incentives for investment in the electricity sector given the dominant role

played by state-controlled enterprises. In the pre-reform era, state plan was very rigid and large amount of resources could be mobilized to develop a few priority

areas of the economy, possibly at the expense of other parts of the economy. Electricity was among the few priority areas. The electricity sector grew very rapidly then, more so than the whole economy did.

After economic reform started, state plan became less rigid and priority was rebalanced. Without given sufficient eco- nomic incentives, the electricity sector would not grow as fast as they once did. This could explain why growth slowed down in the first few years of economic reform.

Since 1985, economic incentives have been gradually introduced. First, investment from outside of the state budget was encouraged and the producers were given more autonomy over electricity generation from the new investment. Then, gov- ernment functions and business activities were separated in the sector, first at the provincial level in 1987 and later at the national level in 1997. With this separation, electricity is no longer considered a government service, albeit with a fee, but a commodity produced and provided by enterprises. This could be one explanation behind a sharp difference between China and India in the electricity sector that electricity stealing is not a significant problem in China and losses in transmission only accounts for about 7% of total output and are mostly due to technical factors. Enterprises that supply electricity in China are subject to less political pressure than they used to be and have strong incentives to enforce payment by consumers.

At the same time, the price formation mechanism has gradually become more market oriented, with the prices mostly reflecting industry average costs and normal rates of return. There are times when the price of electricity is not be allowed to adjust accordingly even though the price of coal used to generate electricity has increased substantially, but this mostly happens during most difficult economic times.

Increased competition is another factor that drives investment. With the 2002 break up of the National Electric Power Corporation, and the establishment of other electricity generating enterprises, competition among them intensified even though they are mostly controlled by the state. Among the big firms, the competition is mainly about who will become

Table 3 Source of funds of investment in urban area by sector: production and supply of electric power and heat power (unit = billion yuan).

Year Total By sources of funds

State budget Domestic loans Foreign investment Self-raised funds Others

2004 499.8 19.2 222.2 18 206.7 33.6 2005 643.7 25.6 296.7 13.6 269.7 38.1 2006 729 31.4 338.7 8.1 304.6 46.3

Sources: China Statistical Yearbook 2005–2007 (NBS, various years).

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 39

the industry leaders and maintain that position. Both of the two power grid companies and seven of the largest electric power generating companies are among the 150 largest state-owned enterprises under the direct control of the State-owned Asset Supervision and Administration Commission (SASAC) of the State Council. All of these state-owned enterprises are un- der strong pressure to be among the top three enterprises in their respective sector or face the risk of being restructured. Furthermore, electricity generating enterprises are currently competing fiercely for most important strategic assets of the sector, plant sites that have convenient access to coal and water, the supply of which is limited. They are worried that if the ideal plant sites are all taken by their competitors, they would face great difficulty expanding their capacity in the future. For all the power generating firms, they have to compete to sell to the monopsonistic buyers of electricity, the power grid companies that are separated along regional lines. This is especially true for smaller power generating firms. If they do not gain enough scale and achieve sufficiently low cost, it is hard for them to survive in the long run. Such competitive pressure also gives firms strong incentives to invest.

Local governments are also an important force behind incentives for investment in the electricity sector. A large electric- ity generating company is at the same time a large tax payer. Sometime, the tax revenue from a large electricity generating company can be a very substantial part of the tax revenue of a county government. There is strong competition among county governments to have the electricity generating company located within their county, even when the region in general enjoys strong geographical advantages in attracting the company. One of the instruments the county government can use is concession on the land price that the county government charges the company. Such competition among the local govern- ments lowers the cost of investment in the sector and encourages investment.

Even though the electric power grid companies face little competition, they still have strong incentives to make invest- ment. In 2007, investment in power grids accounted for 45% of total investment in the electricity sector (SERC, 2008). Similar to electricity generation, the way prices for transmission and distribution are determined is one reason for the strong invest- ment incentives. The prices are determined based on the allowable costs, the allowable rate of return, and the related taxes (NDRC, 2005a). Possible tunneling by the insiders of power grid companies through transfer pricing may also be a contrib- uting factor behind the strong investment drive. All the power grid companies are state owned and the sector is dominated by two power grid companies. However, the provincial branches of the companies enjoy considerable autonomy and many investment decisions are made at the provincial level. Many of the provincial power grid companies have a sister company that is often owned by their employees or managers. The state-owned provincial power grid company does many businesses with the collectively owned sister company, including the procurement of equipment, electricity, and other related services. On paper, the state-owned power grid companies may not be very profitable. However, their collectively owned sister com- panies are often very profitable. Dividend payment from the sister company can be a substantial part of the income of the employees of the state-owned counterpart. Such a relationship between the two companies can create powerful investment incentives for the state-owned power grid company. After all, investment by the power grid company creates demand for equipment and services and is good for the business of the sister company, and this in turn benefits the managers and employees of the power grid company. A consequence of this practice is that the income of some managers and employees of the power grid companies is much higher than that of similar people in non-monopoly sectors. This has becomes a source of resentment against income inequality.

It is possible that these strong investment incentives may give rise to overinvestment. Indeed, there was surplus of supply in 1997 and there have been concerns about overinvestment in the last few years. However, the economy is growing very rapidly and any surplus supply will soon be met by increased demand. The aftermath of the 1997 experience is a case in point. After the government suspended investment in electricity generation for 3 years, shortage of electricity follows a few years later. Of course, the economy will eventually slow down and by then, overinvestment will become more costly. The current system seems to work in the current environment but may not be the first best arrangement in the long run.

The lack of foreign investment in the electricity sector is interesting given that foreign direct investment has played a very significant role in other parts of the economy. Possible reasons are as follows. The transmission and distribution sector is off limit to foreign investment. Even though the power generation sector is open to foreign investment, foreign invested gen- erators are disadvantaged because local relationship is very important for a generator to gain access to the grid, and to secure stable and cheap supply of coal. Furthermore, the most important reason for foreign direct investment in China is not to bring in capital, but to facilitate the transfer of knowhow (Bai, Lu, and Tao, 2009). Technology in electricity generation is easy to transfer even without long term participation of foreign investors, and therefore, the role of foreign investors is not as important as in many other sectors.

2.4.3. Rate of return to investment Using data from the 2007 annual reports of the top 10 publicly listed electricity generating companies, ranked according

to their total asset values in 2007, we estimate the average rate of return on equity of these companies from 2000 to 2007 (Fig. 2). The rates range from 10.45% in 2002 to 12.78% in 2001. These rates are higher than the normal rate of return the government uses to determine the prices, which is 1 percentage point above the annual interest rate paid by long term gov- ernment bonds. The high rates of return on equity are achieved through leverage. They compare favorably with the rate of return to capital of the whole economy net of urban residential housing estimated by Bai et al. (2006).

We also estimate the rate of return to capital in the whole electricity and heat power sector. We estimate the capital stock of the sector using the perpetual inventory approach and the investment data from 1950 to now, and derive the capital in- come data by adding operating profits and production taxes of enterprises. The rate of return is estimated by the ratio of

10.00%

10.50%

11.00%

11.50%

12.00%

12.50%

13.00%

2000 2001 2002 2003 2004 2005 2006 2007

Fig. 2. Average return on equity of the top 10 publicly listed electricity generating companies. Sources: corporate annual reports.

85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 Year

Rate of Return to Capital in the Electricity Sector

0.050

0.100

0.150

0.200

Fig. 3. Rate of return to capital (net of depreciation) in electricity and heat water sector. Sources: authors’ calculation.

40 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

capital income over capital stock at current price (see Appendix A for detailed description of data and estimation method). Fig. 3 shows that the rate of return has been fluctuating around 10% since the last 1990s. Again, these rates compare well with those of the whole economy net of urban residential housing estimated by Bai et al. (2006).

3. Highways

3.1. Statistical summary of the transportation sector

The transportation sector has experienced very rapid growth since 1978. In 1978, the total passenger-km was 174 billion, but in 2006, it was 1919 billion, representing an annual growth rate of 8.9%. Civil aviation grew fastest, from 2.8 billion pas- senger-km to 237 billion passenger-km, or 17.2% a year. Highways follow, from 52 billion to 1013 billion, or 11.2% a year, and then railway, from 109 billion to 662 billion, or 6.6% a year. The passenger-km of waterways dropped from 101 billion to 74 billion, or �1.1% a year. Fig. 4 shows the growth in passenger-km of various modes of transportation with the real GDP growth as the reference.

Since different modes of passenger transportation have very different growth rates, their relative shares in the sector in terms of passenger-kilometer has changed significantly since 1978. Fig. 5 shows the shares in passenger-kilometers of var- ious modes of transportation. It shows that highways have replaced railways as the most important mode of passenger transportation and the importance of civil aviation has increased dramatically, from 1.6% to 12.3%.

Freight transportation has also experienced very rapid growth since 1978. In 1978, the total ton-km was 983 billion, but in 2006, it was 8895 billion, representing an annual growth rate of 8.2%. Civil aviation grew fastest, from 97 million ton-km to 9.4 billion ton-km, or 17.8% a year. Highways follow, from 27.4 billion to 975.4 billion, or 13.6% a year, then waterway, from 378 billion to 5549 billion, or 10.1% a year. The ton-km of railways and petroleum and gas pipelines grew at

(Logarithm with the 1978 Values Normalized to 0)

-1

0

1

2

3

4

5

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

GDP Indices

Railways

Highways

Waterways

Civil Aviation

Fig. 4. Real GDP and passenger-kilometers by various modes of transportation. Sources: China Statistical Yearbook (NBS, various years).

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

P er

ce nt

Railways

Highways

Waterways

Civil Aviation

Fig. 5. Shares in passenger-kilometers of various modes of transportation. Sources: China Statistical Yearbook (NBS, various years).

(Logarithm with the 1978 Values Normalized to 0)

-1

0

1

2

3

4

5

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

GDP Indices Railways Highways Waterways Civil Aviation Petroleum and Gas Pipelines

Fig. 6. Real GDP and freight ton-kilometers by various modes of transportation. Sources: China Statistical Yearbook (NBS, various years).

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 41

approximately the same rate, 5.2% and 5% respectively, with the former going from 535 billion to 2195 billion and the latter going from 43 billion to 166 billion. Fig. 6 shows the growth in ton-km of various modes of transportation with the real GDP growth as the reference.

The relative shares of different modes of freight transportation have also changed significantly since 1978. Fig. 7 shows the shares in ton-km of various modes of freight transportation. It shows that waterways have replaced railways as the most important mode of freight transportation. The importance of highways increased dramatically from 2.8% in 1978 to 14.4% in 1998, but gradually declined to 11% in 2006.

The rapid increase in the volume of transportation is only possible with the improvement of the transportation infrastruc- ture. From 1978 to 2004, the length of highways increased from 890.2 to 1870.7 thousand km. From 1978 to 2006, the lengths of railways in operation, civil aviation routes, and petroleum and gas pipelines increased from 51.7 to 77.1, from 148.9 to 2113.5, and from 8.3 to 48.2 thousand km respectively. Fig. 8 shows the growth of the lengths of transportation routes. The growth in the length of highways is higher than that of railways.

Among highways, expressways have been growing the most rapidly. Their length increased from 100 km in 1988 to 45.3 thousand km in 2006. From 1988 to 2004, the length of all highways grew at an annual rate of 4%, but the length of express- ways grew at a staggering annual rate of 44%. Fig. 9 illustrates this phenomenal growth.

For the railway, we have some data about its technical efficiency. The data shows that technical efficiency has also im- proved together with the increase in the length of railways. Fig. 10 illustrates the trend of some of the indicators of technical efficiency. We can see that the running speed of passenger trains and that of the freight trains have both increased, with the former increasing faster. The abnormality of the 2001 data about freight train speed is probably due to a statistical error. The density of passenger transportation and freight transportation both show significant improvement. The handling time of freight first increased from 1985 to 1999 and has then gradually come down since 1999.

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

P er

ce n

t Railways Highways Waterways Civil Aviation Petroleum and Gas Pipelines

Fig. 7. Shares in ton-kilometers of various modes of transportation. Sources: China Statistical Yearbook (various years).

(Logarithm with the 1978 Values Normalized to 0)

-0.5

0

0.5

1

1.5

2

2.5

3

19 78

19 80

19 82

19 84

19 86

19 88

19 90

19 92

19 94

19 96

19 98

20 00

20 02

20 04

20 06

Year

Railways in Operation

Highways

Civil Aviation Routes

Petroleum and Gas Pipelines

Fig. 8. Length of Transportation Routes. Note: The 2005 and 2006 data for highways use a different definition from that used for earlier years and are therefore left out. Sources: China Statistical Yearbook (various years).

Length of Highways and Expressways (Logarithm with the 1978 Values Normalized to 0)

0

1

2

3

4

5

6

7

19 88

19 89

19 90

19 91

19 92

19 93

19 94

19 95

19 96

19 97

19 98

19 99

20 00

20 01

20 02

20 03

20 04

20 05

20 06

Year

Highways Expressway

Fig. 9. Length of expressway (1988–2006: 100 km to 45.3 thousand km). Sources: China Statistical Yearbook (various years).

Technical Efficiency (Logarithm with the 1978 Values Normalized to 0)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

19 85

19 87

19 89

19 91

19 93

19 95

19 97

19 99

20 01

20 03

20 05

Year

Running Speed of Passenger Trains (km/hr)

Density of Passenger Transportation (10 000 person.km/km)

Running Speed of Freight Trains (km/hr)

Density of Freight Transportation (10 000 tn.km/km)

Handling Time of Freight (hour)

Fig. 10. The technical efficiency of railways. Sources: China Statistical Yearbook (various years).

42 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

In summary, the transportation sector as a whole has grown rapidly. The growth rates of passenger-kilometer and ton- kilometer are not very far below the growth rate of GDP. However, railway has lagged behind. Although its technical effi- ciency has been improving, its growth rates in terms of passenger-kilometer and ton-kilometer are both much lower than that of GDP, and its length has also been growing much slower than that of the other modes of transportation. As a result, its shares in passenger and freight transportation have both declined steadily. In the remainder of this paper, we will discuss the development of the highway and railway sectors respectively.

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 43

3.2. Evolution of highway policy

The evolution of highway policy can be divided into four stages, with more market elements introduced from period to an initial system of complete central planning.

3.2.1. Central planning Before 1978, it was a complete state planning system. The central government was responsible for a few strategically

important national highways, and the local governments were responsible for the construction and management of all the other highways. Roads were considered a pure public service. Investment funds came from state budget in the form of a grant that did not carry any financial cost. Users did not pay any toll, but they paid a monthly highway maintenance fee based their vehicle tonnage. Because of shortage of funds and lack of incentives, the development of highways was slow.

3.2.2. Decentralization In the next stage, state investment funds in local projects were changed from grants to loans in 1980 and the cost of

investment was shifted to the local governments. New fees were introduced in the following years to increase the amount of funds available for highway construction. Among the newly introduced fees was vehicle purchase surcharge, which was started in 1985. The surcharge was 10% of the price for domestically produced vehicles and 15% for imported vehicles. This surcharge goes into the State Highway Construction Fund.

3.2.3. Introduction of tolls and their current regulation Highway tolls were introduced in the third stage. It was first experimented in Guangdong province in 1981 when the local

government borrowed from foreign investors to finance the construction of four bridges along the Guangzhou–Zhuhai high- way and raised funds to construct the Guangzhou–Shenzhen highway. Tolls were collected to secure the financing of these projects. Gradually, toll collection became a widespread practice in the whole country.

Currently, highway tolls are regulated by the Highway Law that was promulgated in 1997 and revised in 1999 and 2004. Among other aspects, the law regulates the types of highway that are eligible for toll, the number of years in which tolls can be collected on a highway, the toll rates, the transfer of the right for toll collection, the setting up of toll gates, and the coor- dination between provinces when the highway crosses the provincial border.

According to the law, only three types of highways are eligible for toll collection. One type includes highways built by local governments above the county level with loans or funds raised from enterprises and individuals. The second type in- cludes the first type of highways of which the right for toll collection has been transferred to an enterprise. The third type includes highways built with investment from enterprises. The first type is called government highways with loans to be paid off (zhengfu huandai gonglu), and the remaining two types are called commercial (jingyingxing) highways. All toll high- ways must have technical grades and scales above government specified thresholds.

The transfer of the right for toll collection has to be approved by the Ministry of Transport for national government toll highways, and the transfer has to be approved by the provincial department of transport and submitted to the Ministry of Transport for record for other toll highways.

The length of the toll period and the toll rates have to be approved by the provincial department of transport. When the right for toll collection of a government highway is transferred, the length of the toll period should be agreed upon by the two sides of the transaction and be approved by the authority that approves the transfer. Toll rates have to be approved by the provincial price regulator as well. The principle governing the determination of toll period and rates is to allow the return of the loans or funds raised for government highways and to allow the return of investment with reasonable profits for com- mercial highways. The toll period is also subject to State Council regulation. According to the State Council Regulation of Toll Roads promulgated in September 2004, the maximum length of toll period is 15 years for government toll roads in general and 20 years for government toll roads in economically less developed central and western regions; the corresponding max- imum lengths for commercial toll roads are 25 and 30 respectively. The law and the regulation were not very clear about when the maximum toll period clock starts ticking when there is a transfer of the right for toll collection. In some cases, the clock starts from zero again after a transfer. There is clearly a coincidence of incentives for both sides of the transfer to agree on a longer period of toll collection. Meanwhile, the two sides of the transfer often have close connection to the government authority that is to approve the terms of transaction; in the case of a government to enterprise transfer, the transferor is the government authority. This loop hole is a subject of much debate.

The setting up of toll gates must have the approval of the provincial government. For highways that cross provincial boundaries, the setting up of toll gates should be agreed upon by all relevant provinces; if an agreement cannot be achieved, the Ministry of Transport will make the decision. The principle is to have unified collection and proportionate sharing of toll fees.

3.2.4. Expansion of financing instruments The institution of toll has paved the way for a wider range of choices of financing instruments. There are several ap-

proaches one can take to finance highway construction in addition to bank loans. One is to form a joint stock company and raise funds in the form of equity investment. Another is to issue bonds. The builder can also choose to sell the right to operate to other investors or even go through an initial public offering after the highway is completed and operated

44 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

for some time and toll income becomes predictable. The fund raised from the transfer of right to operate can be used to pay off bank loans or build new highways. There are also cases in which the builder of the highway raises fund by issuing asset backed securities with future toll income as the underlining asset. An early example is Zhuhai Expressway. In August 1996, Zhuhai City Government registered Zhuhai City Expressway Company in Cayman Islands and used it as a vehicle to issue 200 million US dollars of asset backed securities. The proceeds of the issue were invested to build railway and expressway be- tween the cities of Zhuhai and Guangzhou.

3.3. Investment incentives

3.3.1. Sources of investment According to the 2006 and 2007 Statistical Report for the Development of the Road and Waterway Transportation Sector

published by the Ministry of Transport (various years), total investment in road transport was 623.1 billion in 2006 and 649 billion in 2007. In contrast, total fiscal revenue of all levels of governments was 3876 billion yuan and the total spending on infrastructure from government budget was only 439 billion (NBS, various years); that is, total investment in transport was 16% of all government revenue and 142% of government budgetary spending on infrastructure. Clearly, it is impossible to finance the investment in road transport all by government budget. Most of the funding has to come from other sources.

Table 4 shows the split of investment funds between central budget and local funds, between domestic investment and overseas investment, and between investment made by state-controlled institutions and other firms. According to the Table, the lion share of the investment is made by state-owned enterprises, mostly controlled by local governments, including pub- licly listed companies in which the government has a controlling share. In 2006, 604 out of 648 billion yuan of investment was made by these enterprises.

The records of the 12 publicly listed expressway companies are also consistent with the above observation. Nine of the 12 companies are controlled by a holding company wholly owned by a provincial government. Of the remaining three, North- east Expressway Co. Ltd. is jointly controlled by the provincial governments of Heilongjiang and Jilin; Huabei Expressway Co. Ltd. is controlled by one of the largest state-owned enterprises, China Merchants Group, through its expressway investment arm Huajian Transportation Economic Development Center, with Tianjin, Beijing, and Hebei governments as large sharehold- ers; Shenzhen Expressway Co. Ltd. is controlled by the city government of Shenzhen, even though its largest shareholder is Hong Kong Securities Clearing Company Ltd. Interestingly, China Merchants Group is among the top three shareholders in 8 of the 12 publicly listed expressway companies, holding on average 21% of the equity shares of these companies. The local government owned controlling shareholders are generally under the direct control of the local departments of transport, and their average shareholding in these 12 companies is 42%. Other large shareholders are asset management funds.

We use Jiangsu Expressway Co. which has 27 billion yuan of assets, the largest among all publicly listed expressway com- panies, to illustrate the ownership structure. Its controlling shareholder is Jiangsu Transportation Holding Company, a wholly state-owned enterprise under the provincial government, and it holds 54.44% of Jiangsu Expressway Co., China Mer- chants Group owns 11.69% and other eight largest shareholders are J.P. Morgan Chase & Co., UBS AG, Sumitomo Mitsui Asset Management Limited, HSBC Halbis Partners Limited, Schroder Investment Management Limited, and three securities com- panies in China.

According to the Statistical Report for the Development of the Road and Waterway Transportation Sector published by the Ministry of Transport (various years), the most important source of the investment in road transport is bank loan, which ac- counts for 40.7% of the total in 2006. The loans are backed by government guarantees and toll revenue. The second most important source is fund raised by the local government or enterprises controlled by them, which accounts for 32.8% of the total in the same year. Central government funding is small, and is mainly used to subsidize the construction of national trunk roads, road construction related to poverty reduction and national defense. It comes from central government bond and vehicle purchase tax. Few intercity highways are built entirely with funding from government budget.

3.3.2. Incentives and problems How can these enterprises yield sufficient returns to their investment so that investment in road transport is sustainable?

One of the most important reasons is highway toll and the regulations about toll. As long as the highway is not built only with government budget and satisfies some technical conditions, the operator is entitled to collect tolls on the highway.

Table 4 investment in fixed assets in urban area by jurisdiction of management and registration status: road transport (unit = billion yuan).

Year Total By jurisdiction By registration status By registration status

Central Local Domestic HK, Macao, Taiwan Foreign State Collective Private

2004 466.55 20.07 446.48 462.65 2.69 1.21 444.21 3.33 1.98 2005 558.14 28.44 529.70 550.58 4.98 2.58 520.31 4.21 4.22 2006 648.16 16.40 631.77 640.85 5.34 1.98 604.35 14.91 24.35

Sources: China Statistical Yearbook 2005–2007 (NBS, various years).

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 45

The principle governing toll collection is that the operator can recover the cost of investment. In the case where the operator is an enterprise, the regulation on tolls also guarantees a reasonable rate of return in principle.

The commitment of the regulation to guarantee the recovery of investment cost or reasonable return to investment is credible because most of the operators of toll highways have strong connections to the regulator. The more likely problem is regulatory capture by the operator, rather than lack of commitment from the regulator. One example of capture is extend- ing the toll period beyond what is allowed by the State Council regulation using loopholes therein regarding the transfer of toll rights that we discussed above. Such an environment may not be the first best for public interests, but in the case where demand is strong and willingness to pay by the users is high, the efficiency loss from capture may be lower than that from lack of commitment.

Dealing with toll regulation is not the only area in which government connection helps. We again use Jiangsu Expressway Co. as an example to illustrate areas in which government connection is important in the highway business. One is project approval and toll regulation. When the company wanted to upgrade the Shanghai–Nanjing Expressway in 2004, it needed the approval of National Development and Reform Commission of the central government. To recoup the cost of the upgrade, it needed the approval of the provincial government to extend the toll period for 5 years beyond the original toll period. When they wanted to upgrade the Shanghai–Nanjing grade-two highway to grade-one highway, they again needed the ap- proval of the provincial government to extend the toll period by 12 years. The second is tax concession and land concession. Companies can get tax exemption, reduction, or rebate when they invest in key basic infrastructure. However, these conces- sions have to be negotiated with the Department of Finance of the province and approved by the Ministry of Finance of the central government. The situation with land concession is similar. The third is government help with commercial financing. To get bank loans, the company needs loan guarantee from qualified entities. In the case of Jiangsu Expressway Co. loan guar- antee has been provided by its largest shareholder, Jiangsu Transportation Holding Company, a wholly state-owned enter- prise. Government connection has also been crucial for the company to get approval for private placement of its shares before its IPO and its listing on Hong Kong Stock Exchange and Shanghai Stock Exchange. The last item is future investment opportunities. Excluding those held by Jiangsu Expressway Co. Jiangsu Transportation Holding Company still holds 2016 km of expressway in its portfolio. The rights to operate these expressways are potential investment opportunities for Jiangsu Expressway Co. Similar to getting favorable terms of toll collection, other benefits from government connection discussed here also promotes investment.

The importance of government connection is not without its costs. One result is that there are very few private firms engaging in the highway business and another is that there is very little foreign direct investment in the sector, as Table 4 shows. Another barrier for private firms to enter the business is that a highway company has to have at least 30% equity to be eligible for bank loans. Given the policy risks involved with highway projects, it is very hard for private investors to raise such fund.

The dominant role played by the government in highway development comes with high political costs. Corruption is ram- pant in this sector. Since 1997, 20 director generals or deputy director generals of various provincial departments of trans- port have been convicted of bribery. In the case of Henan province, three consecutive director generals have been convicted of the crime one after the other in unrelated cases. The provincial departments of transport have almost unchecked power in offering highway construction contracts, even though the contracts are ostensibly offered by a tender process. In some cases, the officials leaked the reservation price to some bidders. The officials also control a large sum of highway construction funds and can allocate the funds to lower level highway authorities with much discretion. The reason for the high incidence of cor- ruption in this sector is not lack of penalty. Of the twenty cases above, there are five death penalties. The temptation is so strong that even the risk of death penalty cannot deter corruption. Table 5 presents a list of bribery cases involving top pro- vincial transport officials.

3.3.3. Rates of returns to investment To quantify investment incentives, we estimate the rate of return to investment. Using data from the annual reports of the

12 publicly listed expressway companies, we find that the average rate of return on equity is 10.5%, with an average leverage ratio of 40%.

We also try to estimate the rate of return to investment for the whole sector. The National Bureau of Statistics of China provides data on investment and the breakdown of value added for the transport, storage and postal service sector. This in- cludes all modes of transportation plus storage and postal service. We use two methods to estimate the rate of return in this sector. One method is similar to the one used in Bai et al. (2006). This method deducts economic depreciation from capital income to get the net rate of return. The other method uses accounting depreciation. More detailed description of data and estimation method is given in Appendix A. The estimated rates of return to capital are given in Fig. 11. The rates of return in the transport, storage, and postal service sector are very respectable; they compare favorably with the rate of return to cap- ital of the whole economy net of urban residential housing estimated by Bai et al. (2006).

Fig. 12 shows the composition of fixed asset investment in the transport and postal service sector in 2006. Investments in highway transport and railway transport together account for 76% of the total. We will show later that the rate of return to capital in the railway sector is much lower than the estimated rate of return in the whole transport and postal service sector. Therefore, we can speculate with some confidence that the rate of return to investment in the highway transport sector should give the investors strong incentives to invest.

Table 5 Top provincial transport officials convicted of bribery.

Year Province Rank of the official Penalty

1997 Henan Director general 15 years 2000 Sichuan Director general Death, commuted 2000 Sichuan Deputy director general Death 2001 Henan Director general Life 2001 Hunan Deputy director general Life 2002 Guangxi Deputy director general 11 years 2002 Guangdong Deputy director general 13 years 2003 Guizhou Director general 17.5 years 2004 Guizhou Director general Death 2004 Heilongjiang Deputy director general Life 2004 Yunnan Deputy director general 2.5 years 2005 Guangdong Director general 13 years 2005 Beijing Deputy director general Death, commuted 2006 Henan Director general Life 2006 Jiangsu Director general 20 years 2006 Anhui Director general 10 years 2006 Hebei Deputy director general 14 years 2007 Zhejiang Director general Life 2008 Fujian Deputy director general Death, commuted

Xinjiang Director general No news of conviction

Source: collected by authors.

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06

Year

net of accounting deprecition net of economic deprecition

Fig. 11. Rates of return to capital: transport, storage, and postal service. Sources: authors’ calculation, see Appendix A for details.

58%

18%

9%

7% 4%

3%

1%

Highway

Railway

Waterway

Urban Public

Aviation

Storage

Others

Fig. 12. Composition of fixed asset investment in transport and postal service sector. Sources: China Statistical Yearbook (NBS, 2007).

46 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

4. Railway

4.1. Evolution of policy

4.1.1. 1949–1982: rigid planning and low prices The Ministry of Railway was part of the rigid planning system. There was no separation of government functions with

business activities. The Ministry of Railway had very little autonomy and the local railway bureaus had even less. Prices were uniformly set by the central authority. In 1955, the price was 0.0165 yuan per ton-kilometer for freight and 0.0149 yuan per person-kilometer for passengers. By 1967, the freight price had dropped to 0.01438 yuan per ton-kilometer.

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 47

4.1.2. 1982–1985: greater local autonomy and some profit retention by the railway system Some autonomy on planning, finance, employment, personnel decisions was given to the local bureaus of railway. At the

same time, the railway system could retain some profit according to a formula after tax payment. Railway service prices were increased but they still lagged behind the general price level. Compared to 1955, the 1986 freight price dropped by 20.8% relative to the consumer price index and passenger price by 15.7%.

4.1.3. 1986–1992: financial independence of the railway system In March 1986, the State Council gave the railway system greater financial autonomy and responsibility. The railway sys-

tem was exempt from income tax and a few other taxes. The tax revenue and profit that were previously paid to the State Treasury were now retained by the Ministry of Railway to invest in the railway system. The railway system was separated from state finance but the railway system itself was still the same entity that combined government functions with business activities. In this period, prices were increased by very large margins.

4.1.4. 1993–2002: corporatization In 1993, the railway system started experimenting with corporatization. Guangdong railway bureau was restructured

into a corporate group in February, 1993. Then Dalian railway bureau became Dalian Railway Limited in 1995. Guangz- hou–Shenzhou railway was publicly listed in Hong Kong and New York in 1996. The whole system went through corpora- tization in 1999. Five non-core businesses including engineering, construction, equipment, materials, and communications were separated from the Ministry of Railway and became independent enterprises in 2000. The corporatization effort at- tempted to separate government functions from business activities in the system. In the same period, the price policy be- came more flexible. Some variation is allowed in prices. Prices were allowed to depend on the quality, route, and season of the services.

4.1.5. 2002 – now: more price flexibility The National Planning Commission issued a document in 2002 that allowed a lot more flexibility in railway prices. The

document referred railway prices as under government guidance rather than determination. The document specified ranges for price adjustment, instead of exact prices. Within the range, the Ministry of Railway has the authority to approve the application of price adjustment submitted by local railway bureaus. Public hearings are also conducted before main price adjustments are made.

4.2. Investment

As shown in Fig. 13, fixed asset investment in the railway sector went through a few steps. It was more or less stable in each of three periods: at around 20 billion yuan a year between 1991–1992, 57 billion yuan a year between 1993–1997, and 87 billion yuan a year between 1998 and 2004. However, it has been increasing very rapidly since 2004. The growth rates are 51%, 52%, and 22% respectively in 2005, 2006, and 2007. One possible reason for the recent rapid growth is the adoption of the ‘‘medium and long term plans for railway network” by the state council in 2004 (SCC, 2004). The plan includes a target growth of 39% in the total length of railways from 2002 to 2020. The target was recently revised up to 67% (SCC, 2007). From Fig. 14, we can see that more than 80% of the investment in basic construction in the sector has been made by the ministry of railway. Table 6 shows that almost all the fixed asset investment is made by domestic state-controlled organizations, and around 90% of the investment is made by the central government. Table 7 shows that state budget plays the most important in fixed asset investment in the railway sector. All these observations are quite different from the electricity and highway sectors.

0

50

100

150

200

250

300

1991 1993 1995 1997 1999 2001 2003 2005 2007

Year

Fixed Asset Investment in Railway (unit = billion yuan)

Capital Construction Improvement Vehicle Purchase

Fig. 13. Fixed asset investment in the railway sector: 1991–2007. Sources: Yearbook of China Transportation and Communications (CCCTCY, various years); Statistical Communiqué of Ministry of Railway (MOR, various years).

82.0%

84.0%

86.0%

88.0%

90.0%

92.0%

94.0%

96.0%

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Year

Fig. 14. Share of capital construction investment made by the ministry of railway. Sources: Statistical Communiqué of Ministry of Railway, 1998–2007 (MOR, various years).

Table 6 Investment in fixed assets in urban area by jurisdiction of management and registration status: railway (unit = billion yuan).

Year Total By jurisdiction By registration status By registration status

Central Local Domestic HK, Macao, Taiwan Foreign State Collective Private

2004 84.63 75.27 9.36 84.63 0.00 0.00 83.78 0.42 0.04 2005 126.77 112.18 14.59 126.72 0.00 0.04 124.45 0.99 0.05 2006 196.65 176.55 20.11 196.56 0.03 0.07 191.79 3.61 1.16

Sources: China Statistical Yearbook, 2005–2007 (NBS, various years).

Table 7 Source of funds of investment in urban area by sector: railway (unit = billion yuan).

Year Total By sources of funds

State budget domestic loans Foreign investment Self-raised funds Others

2004 82.95 32.27 17.85 0.47 25.26 7.10 2005 128.84 44.62 33.37 3.03 38.84 8.98 2006 194.56 58.76 33.15 2.49 71.05 29.10

Sources: China Statistical Yearbook, 2005–2007 (NBS, various years).

48 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

From Table 8, we can see that the railway system has been profitable since 1998. However, the rate of return to invest- ment in railway is low. Fig. 15 shows the rate of return to capital since 1991. The rates are low in recent years even when taxes are included in capital income. Given these low rates, it is hard to imagine profit-oriented enterprises would have any incentives to investment in the railway sector.

5. Conclusion

Of the three sectors we consider in this paper, railway stands out being very different from the electricity and highways sectors. The electricity sector, both in terms of installed capacity and output, has kept in pace with the growth of the econ- omy, and so has highway transport in terms of business volume. The growth in the length of expressways has also been phe- nomenal. The railway sector, however, has seen its business volume growing much slower than GDP for both passenger and freight, and its share in the transport market shrinking fast. Its route length has grown at the lowest rate among all major modes of transportation.

One major difference between the railway sector and the other two sectors is that the former is still very centralized and its business is still essentially run by the central government, specifically the Ministry of Railway. Almost all of the invest- ment is made by the ministry. In contrast, in the electricity sector and the highway sector, there is much autonomy at the provincial level and at the enterprise level. Enterprises in these two sectors have very strong profit orientation and many of them are publicly listed companies, even though the government often maintains a controlling share of them. Local govern- ments are also motivated to help enterprises in the two sectors. In the electricity sector, local governments compete for investment in their region to increase their tax base, by offering land concession among others. In the highway sector, local governments are often one of the owners of the highway companies and their incentives to help the companies are also very strong. Competition among enterprises created by decentralization is another factor that drives investment in the electricity sector.

Prices in the railway sector are yet to be determined on commercial basis, despite a general trend of relaxation of price rigidity. Other goals dominate commercial viability in determining the prices. There are clear indications that prices are too low for efficiency. According to the Development Research Center of the State Council (2005), only 30–40% of the demand for railway freight transport is met. The shortage of railway transport supply is especially severe for the transport of coal. It is

Table 8 Financial indicators of the railway system (1991–2007).

Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Fixed Asset Investment

16.7 22.8 44.9 56.5 57.3 61.3 65.6 83.1 81.6 80.5 89.8 96.4 86.1 90.2 136.4 207.1 252

Capital Construction 8.6 12 26.7 36.3 36.6 36.8 38.8 56.2 56.6 52.7 58.1 62.4 52.9 53.2 88 154.3 177.2 Improvement 4.1 6.5 10.3 9.6 10 12.3 14.3 15 15.2 14.3 16 16.4 16.4 19.2 21.8 20.9 19.5 Vehicles 4 4.3 7.9 10.6 10.7 12.2 12.5 11.9 9.8 13.5 15.7 17.6 16.8 17.8 26.6 31.9 55.3 Total revenue 44.5 47.9 54.7 58.3 63.2 75.6 84.2 92.2 99.9 109.7 134.6 142 148.3 179.4 201.9 236.4 Passenger 12.2 13.8 15.9 17.3 20.2 23.7 26.1 29.1 33.2 36.9 46.2 49.7 47.7 59.3 63.8 72.8 Freight 28.7 30.1 31.6 34.6 35.8 40.6 46.1 50.3 52.2 56.6 68.9 70.5 78.4 94.4 110.6 128.1 Others 3.5 4 7.3 6.4 7.3 11.3 12 12.8 14.5 16.2 19.6 21.9 22.2 25.8 27.5 35.6 Transportation costs 29.4 35.8 47.6 57.1 64.1 70.5 81.2 84.8 90.3 100.5 123.4 131.1 139 165.7 182.6 Depreciation 12.1 13.5 14.9 6.6 7.4 9.7 11.3 12.4 31.8 16 19.8 22.3 20.2 23.6 26.2 Other costs 2.9 3.6 3.4 2.9 3.3 4 4.4 5 4.9 4.5 6 5.8 6 7 10.6 11.9 Transportion profits 9.9 5.9 1.3 �2.9 �6.4 �1.4 �2.6 0.03 2.7 3.4 2.1 2.5 1.8 4.9 8.6 6.9 Taxes paid 2.5 2.7 5.4 3 3.2 3.6 2.8 3 3.4 3.8 4.6 5.9 4.8 7.3 9.4 10.3 Gross value of fixed

capital 141.9 152.9 170.7 332.1 361.4 417.1 449.9 494.7 541.5 526.9 640 713.3 787.2 865.6 929.4 1055

Net value of fixed capital

101.1 107.2 119.6 219 240.9 288.2 311.4 345.8 381.8 383 458.4 515 565.8 628.8 666.8 769.7

Sources: Year Book of China Transportation and Communications (CCCTCY, various years); Statistical Communiqué of Ministry of Railway (MOR, various years).

-0.050

0.000

0.050

0.100

0.150

19 91

19 93

19 95

19 97

19 99

20 01

20 03

20 05

Year

Net of Taxes Gross of Taxes

Fig. 15. Rate of return to capital of the railway system (1991–2006). Sources: Authors’ calculation based on data from Year Book of China Transportation and Communications (CCCTCY, various years).

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 49

sometimes difficult for passengers to get tickets too. As a result of the low prices, the rate of return to investment in the rail- way sector is very low and it is impossible to attract profit-oriented investors in the sector. In contrast, prices in the other two sectors are mostly determined according the principle that the recovery of investment cost and sometimes reasonable rate of return to investment are guaranteed. As a result, the rate of return to investment is sufficiently high to attract invest- ment in a sustainable manner. Of course, there is also risk that prices in these sectors become too low due to political pres- sure. Such pressure may affect investment incentives.

Observations suggest that investment in the railway sector is too low. For example, even though there is clear advantage of railway transportation of coal over highway transportation of coal, the latter is still very common. About 25% of coal is transported via highway out of Shanxi province, the largest coal producing province in China. The province is investing 28.8 billion yuan to build more highways for coal transportation. There is strong reason to believe that efficiency should im- prove if the capacity of the railway system increases to accommodate such need for the transportation of coal. Recent surge in railway investment seems a welcome trend, but the new investment is still made mostly by the Ministry of Railway, and the jury is still out on whether it will be allocated to areas where it is most needed.

The other two sectors are not without their problems. In the highway sector, bureaucrats still have too much control over the planning and approval of projects, the provision of loan guarantee, the choice of contractors, the approval of toll terms, and the approval of the transfer of toll rights. The relationship between the bureaucrats and the enterprises in the sector is too close. The resulting corruption problem is politically very costly. It may also cause economic inefficiency. For example, it may lead to poor quality and high cost of construction when highway construction contracts are offered to the cronies of the bureaucrats in charge. Regulatory capture resulting from the close connection between the government and the enterprises increases the cost of usage to the consumers. In the electricity sector, some companies have strong market power. Companies with strong connections to the government or have monopoly power do very well financially. As a result, there is not a level playing field for other firms to compete in. Employees in both sectors tend to be compensated much better than employees in other sectors. This increases income inequality and political resentment.

In the short run, some of these problems may be part of the solution. For example, the close connection of the firms to the government in the two sectors may help prevent the regulators from arbitrarily reducing the promised price after invest-

50 C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51

ment has been made. Rent extraction opportunities that are associated with investment in the sectors may cause overinvest- ment. Naturally, overinvestment is inefficient. However, given the high growth rate of the economy, any overcapacity will soon be filled by newly developed demand and the efficiency loss can be a tolerable sacrifice in exchange for rapid expansion of the sector. In sum, some of these practices may serve as useful transitional arrangements that solve problems in the short run, even though they are not the first best solution in the long run.

Acknowledgment

We thank the editor and an anonymous referee for their comments and suggestions, China National Science Foundation for financial support (grant number 70625002), and Zijian Cheng, Qi Liu, Weiqing Luo, Zhenjie Qian, Changzheng Zhao, and Luan Zhao for their excellent research assistance.

Appendix A. The estimate of return to capital

A.1. Electricity and heat power

Capital income is proxied by operating profits and production taxes of enterprises with independent financial accounts before 1998 and of SOEs and non-SOEs with annual sales over 5 million yuan since 1998 in electricity and heat water indus- try from various issues of China Statistical Yearbook.

Capital stock at current price in electricity and heat water is estimated by perpetual inventory method. The procedure is as follows: (1) proxy investment in fixed assets in electricity industry with that of SOEs before 1981 (source: DOI-NBS, 2002) and the sum of capital construction and capital improvement in electricity and heat water since 1981 (source: China Statistial Yearbook, NBS, various years); (2) use this proxy and national investment in fixed assets to estimate the share of electricity in national fixed capital formation; (3) estimate fixed capital formation in electricity industry with the share obtained in (2) and national fixed capital formation (source: DNA-NBS, 2007); (4) use the composition of national investment in fixed assets (source: China Statistial Yearbook, NBS, various years) to separate the estimate in (3) into construction investment and equipment investment; (5) price indices of construction investment and equipment investment for 1978–1989 are proxied by prices indices of value added in construction industry (source: DNA-NBS, 2007) and PPI in machine industry (source: China Statistial Yearbook, NBS, various years); price indices of construction investment and equipment investment for 1990–2006 are taken from China Statistical Yearbook (NBS, various years); (6) estimate capital stock of electricity industry at 1978 price by type of investment with price index for each type; (7) estimate capital stock of electricity industry at current price by type of investment with price index for each type; (8) add up capital stock at current price by type of investment to get capital stock at current price in electricity industry. Please refer to Bai et al. (2006) for detailed explanation of this method of esti- mating the capital stock.

Capital returns net of accounting depreciation are estimated by the ratio of capital income over capital stock at current price.

A.2. Transport, storage, and postal service

The breakdown of value added in the sector is from DNA-NBS (2007). The investment data are estimated with capital con- struction in this sector before 1980. Those in other years are similar to those in steps (2) to (7) for the electricity and heat water sector.

Economic depreciation rate is estimated by the weighted average of depreciation rate in construction investment and equipment investment, using capital stock by the two types of capital stock at current price as weights.

References

Aschauer, David Allen, 1989. Why is infrastructure important? Journal of Monetary Economics 23, 177–200. Bai, Chong-En, Lu, Jiangyong, Tao, Zhigang, 2009. Capital or Knowhow: The Role of Foreign Multinationals in Sino-Foreign Joint Ventures. Tsinghua

University, Mimeo. Bai, Chong-En, Hsieh, Chang-Tai, Qian, Yingyi, 2006. Returns to capital in China. Brookings Papers on Economic Activity 59 (2), 61–101. Compilation Committee of China Transportation & Communications Yearbook (CCCTCY), various years. Yearbook of China Transportation and

Communications. Press of China Transportation & Communications Yearbook, Beijing. Compilation Committee of China Electricity Yearbook (CCOCEY), various years. China Electricity Yearbook. Press of China Electricity Yearbook, Beijing. Demurger, Sylvie, 2001. Infrastructure development and economic growth: an explanation for regional disparities in China? Journal of Comparative

Economics 29 (1), 95–117. Department of Investment of National Bureau of Statistics (DOI-NBS), 2002. Statistics on Investment in Fixed Assets of China. Statistical Press, Beijing. Department of National Accounts of National Bureau of Statistics (DNA-NBS), 2007. Data of Gross Domestic Product of China: 1952–2004. Statistical Press,

Beijing. Development Research Center of the State Council. Reform of the Railway System, 2005. <http://www.drcnet.com.cn/DRCNet.Channel.Web/subject/

tltzgg.asp>. Fan, Shenggen, Zhang, Xiaobo, 2004. Infrastructure and regional economic development in rural China. China Economic Review 15 (2), 203–214. Haughwout, Andrew F., 2002. Public infrastructure investments, productivity and welfare in fixed geographic areas. Journal of Public Economics 83 (3),

405–428. Holtz-Eakin, Douglas, 1994. Public-sector capital and the productivity puzzle. The Review of Economics and Statistics 76 (1), 12–21.

C.-E. Bai, Y. Qian / Journal of Comparative Economics 38 (2010) 34–51 51

Hulten, Charles R., Schwab, Robert M., 1991. Is There Too Little Public Capital? Infrastructure and Economic Growth. American Enterprise Institute Discussion Paper, Feburary.

Kavanagh, Catherine, 1997. Public capital and private sector productivity in Ireland, 1958–1990. Journal of Economic Studies 24 (1/2), 72. Keeler, Theodore E, Ying, John S., 1988. Measuring the benefits of a large public investment: the case of the US federal-aid highway system. Journal of Public

Economics 36 (1), 69–85. Li, Zhigang, 2005. Measuring the Social Return to Infrastructure Investments Using Interregional Price Gaps: A Natural Experiment. The University of Hong

Kong, Mimeo. Lynde, Catherine, Richmond, James, 1992. The role of public capital in production. The Review of Economics and Statistics 74 (1), 37–44. Ministry of Railway (MOR), various years. Statistical Communique of Ministry of Railway, PRC (in Chinese, Tielu Tongji Gongbao). Ministry of Transport, various years. Statistical Report for the Development of the Road and Waterway Transportation Sector (in Chinese, Gonglu Shuilu

Jiaotong Hangye Fazhan Tongji Gongbao). Morrison, Catherine J., Schwartz, Amy Ellen, 1996. State infrastructure and productive performance. American Economic Review 86 (5), 1095–1111. Munnell, Alicia H., 1990. Why has productivity growth declined? productivity and public investment. New England Economic Review 59 (January/

February), 3–22. Munnell, Alicia.H., 1992. Policy watch: infrastructure investment and economic growth. The Journal of Economic Perspectives 6 (4), 189–198. National Bureau of Statistics (NBS), various years. China Statistical Yearbook. Statistical Press, Beijing. National Development and Reform Commission (NDRC), 2005a. Interim Provisions for the Administration of Grid Power Prices, vol. 514. NDRC Pricing (in

Chinese, Shangwang Dianjia Guanli Zanxing Banfa). National Development and Reform Commission (NDRC), 2005b. Interim Provisions for the Administration of Power Selling Prices, vol. 514. NDRC Pricing (in

Chinese, Xiaoshou Dianjia Guanli Zanxing Banfa). National Development and Reform Commission (NDRC), 2005c. Interim Provisions for the Administration of Transmission and Distribution Prices, vol. 514.

NDRC Pricing (in Chinese, Shupei Dianjia Guanli Zanxing Banfa). Rubin, Laura S., 1991. ‘‘Productivity and the Public Capital Stock: Another Look.” Federal Reserve Board Discussion Paper. Shirley, Chad, Winston, Clifford, 2004. Firm inventory behavior and the returns from highway infrastructure investments. Journal of Urban Economics 55

(2), 398–415. State Council of China (SCC), 1985. Interim Provisions on Encouraging Fund Raising Production and Executing Multiple Prices in Electricity Sector, Ratified

and Forwarded by SCC and Proposed by National Economic Council, State Planning Commission, Ministry of Water and Electricity, and State Price Control Bureau (in Chinese, Guanyu Guli Jizi Bandian He Shixing Duozhong Dianjia De Zanxing Guiding).

State Council of China (SCC), 2002. System Reform Plan for the Electricity Sector. State Council of China (SCC), 2004. Medium and Long Term Plans for Railway Network Januarary 7th, (in Chinese, Zhongchangqi Tieluwang Guihua). State Council of China (SCC), 2007. Medium and Long Term Plans for Comprehensive Transportation Network, Passed by the 195th Standing Meeting,

October, (in Chinese, Zonghe Jiaotongwang Zhongchangqi Fazhan Guihua). State Electricity Regulatory Commission (SERC), 2008. Electricity Regulation Annual Report 2007, April, (in Chinese, Dianli Jianguan Baogao). Tatom, John A., 1991. Public capital and private sector performance. Federal Reserve Bank of St. Louis Review 73 (3), 3–15.