Abstract

We study empirically credit availability of listed firms in China for the years 2003–11 to uncover underlying trends in economic policies. The estimations indicate increasing favoritism of state-owned firms in credit availability, consistent with a drive toward “state capitalism.” Initially, favoritism applied mainly to firms owned by the central government, but the difference between central and local government firms gradually diminished to insignificance. These results signal that economic policies pushed the Chinese economy from the path toward a market economy and state capitalism, and that the economic importance of local governments was growing.

1.  Introduction

Since the “reform and opening” policies were launched in 1978, China has witnessed decades of strong development of the private corporate sector and a reduction in the role of state-owned enterprises (SOEs). This process toward a market economy based increasingly on private ownership of the means of production has contributed to remarkable growth in the economic efficiency of Chinese enterprises, and the economic welfare of its citizens. In spite of the official policy line of the country's unwavering commitment to “socialism with Chinese characteristics,” the accession of China into the WTO in 2001 raised expectations of strengthening of market oriented economic policies (Bajona and Chu 2010).

A growing body of evidence indicates, however, that economic policies during Hu Jintao's government (2003–13) were not aimed at continuing China's reform into a market economy. Important legislative reforms, such as reform of competition law, financial regulation, and capital account regulation, were halted or put on a slow track. According to critics, economic policies during this “lost decade” of economic reform increasingly favored state-owned firms, thereby promoting “state capitalism” instead of a private market economy (Walter and Howie 2011). Due to the opaqueness of the Chinese political system, however, claims about a push toward state capitalism are not easy to verify.

In the academic literature, among the first to present rigorous empirical evidence of such a policy shift was Zhao (2009). Based on accounting data, he found a marked downturn in the fortunes of private firms in China around 2003. Further building on survey evidence, he linked this phenomenon with discriminatory government policies against private firms, which hampered their credit availability. Credit availability is arguably a good indirect indicator of underlying policies in the opaque Chinese political system, because banks’ credit policies are strongly influenced by underlying political trends.

Discrimination against private firms in access to credit has since been confirmed in many empirical studies using alternative approaches (Ding, Guariglia, and Knight 2013; Guariglia, Liu, and Song 2011; Poncet, Steingress, and Vandenbussche 2010; Héricourt and Poncet 2009). The econometric findings are further strengthened by recent macroeconomic indicators that show that the role of SOEs is no longer declining in China (OECD 2013).1 In international comparisons, China stands out from its peers by the large size of its SOE sector (Kowalski et al. 2013).

Our study of credit availability of listed firms in China between 2003 and 2011 contributes to this debate with a focus on how the bias in favor of state-owned firms’ access to credit developed during the previous government's reign. The estimations are based on the novel methodology by Herrala (2014). They show that favoritism of state-owned firms in access to credit grew continuously more pronounced until at least 2011, that is, even after the accommodative policies in response to the first phase of the 2008–09 global financial crisis had abated. This result rebukes the often expressed response that favoritism of state-owned firms was a temporary measure that emerged as a reaction to the international financial crisis. That favoritism of state-owned firms increased rather than decreased after cyclical normalization of economic policies had taken place suggests that the issue is structural rather than cyclical in nature.

To shed further light on the nature of favoritism of state-owned firms, we also test whether credit availability of local government–owned firms corresponded with that of state-owned firms in general. An intriguing novel finding is that favoritism of local government–owned firms shows different dynamics than that of state-owned firms in general, reflecting changes in the role of local governments during the past decade. The estimations show that at the start of Hu Jintao's regime, firms owned by the central government in particular enjoyed favoritism in credit availability. In contrast, the credit availability of local government firms was approximately at par with that of private listed firms at that time. This result is consistent with the view that initially favoritism of state-owned firms reflected a drive toward centralized state capitalism: promotion of corporate giants under the control of the central government to make them competitive both internationally and domestically.

Nevertheless, the estimations indicate that the difference in credit availability between local government–owned firms and state-owned firms in general vanished during the latter part of the past decade. This result may signal a change in the economic responsibilities and power of local governments: Although local governments have always been instrumental in policy implementation in China, the economic decisions that they make have grown in importance and complexity. This process of “localization” of state capitalism accelerated during the latter part of the past decade when, faced with growing demands for improving their own economies, local governments responded by harnessing local corporations as a vehicle to achieve their goals. As a result, the balance of economic decision making shifted from the central government to the local level.

In the next section, we set the stage with a discussion of economic policies in China more generally during the period of interest. The empirical methodology to test favoritism of state-owned firms in credit availability is formalized and the estimation data is discussed in the following sections. This is followed by the estimation results of credit availability and a robustness analysis. A summary and our views, along with a discussion and suggestions for a future agenda, conclude.

2.  General observations about economic policies during President Hu Jintao's regime

The 10th five-year plan for years 2001–05 laid down the main tasks of the government during the first term of President Hu Jintao's government. It called for continued high growth and stable inflation, the strengthening of the industrial sector's international competitiveness, and the industrial sector's technological advancement. Growing environmental concerns and the development disparity between the eastern and other parts of China were also high on the agenda.

The 11th five-year plan for years 2006–11 restated high growth targets and reflected ever-growing environmental concerns as well as the need to strengthen the services sector. During its two terms in power, Hu Jintao's government succeeded in continuing the rapid growth of the Chinese economy: Real GDP grew by about 10 percent per annum on average, exceeding the official growth targets. The regime also succeeded in maintaining near-stable consumer prices, which grew at 2 percent per annum on average.

The role of local governments has always been important in policy implementation in China. Nevertheless, many observers have noted a marked increase in the complexity and economic significance of the issues faced by local authorities during Hu Jintao's government (OECD 2013). Calls for public service equalization across China as well as demands for high economic growth even during the international economic crisis placed severe fiscal strains at the local level. At the same time, the local governments were afforded limited fiscal options, because their right to issue debt and raise tax revenue was severely restricted by law. To overcome this challenge, local governments had to devise ways to step out of the fiscal straightjacket.

Local governments responded to these pressures by setting up special institutions outside the traditional fiscal economy to foster local economic development. Typically, the special institutions would secure financing from local banks based on guarantees or collateral provided by the local authorities, to use in the local development projects. The amount of funds that flowed through these special institutions is not known, but estimations by international organizations suggest these to have been considerable, possibly close to one-third of the Chinese GDP at the end of the last decade (IMF 2012).

These developments have been followed by heated debate about the role of local governments in the Chinese economic model. Whether intended or not, policymakers face a new reality where local governments have considerable economic and financial commitments and responsibilities. Many call for a reform of the tax system to better reflect the increased weight of local governments in economic decision making.

Parallel to these developments, reform of the financial system proceeded slowly. Around 2006, a number of regulatory changes widened the scope of firms listing in Shanghai and Shenzhen, as well as the scope of domestic and foreign investors investing in these markets. These developments were accompanied by a marked increase in the number of listings, and a boom in stock prices.

Much of the financial system remained under tight government control. Throughout the period of study, the financial system was dominated by domestic banks, which the policymakers controlled with an array of levers significantly broader than that in developed countries. The government controlled the reference rates for lending and deposits, reserve requirements, and supervisory parameters, and it was the majority owner of the four largest banking institutions, which still account for 70 percent of commercial banks’ aggregate balance sheets. The government also had other channels of influence over banks’ credit policy, for example, via party-exerted influence over bank executives. Throughout the period, the government imposed quantitative targets for loan and money growth. The effectiveness of window guidance was amplified by the low degree of development of the corporate bond market, which rendered even large corporations dependent on bank lending.

3.  Methodology

To uncover favoritism towards state-owned firms, we study credit availability of listed firms. Because credit availability is strongly influenced by political trends rather than market forces in China, credit availability is a promising indirect indicator of underlying policies in the opaque political system. Credit availability is influenced by the political processes via many channels, including the “window guidance system” (political steering over banks’ credit policies by the central bank), state ownership of the largest banks, and membership of bank directors in the ruling communist party (see the previous section for more discussion). Moreover, because banks assess the economic prospects of their client firms as part of credit risk evaluation, changes in credit availability mirror banks’ views about the underlying political trends in China. If banks observe a political drive toward increased favoritism of specific types of firms, then they are likely to respond by increasing credit availability to such firms.

We quantify credit availability of firms by stochastic frontier analysis of a firm-level sample.2 Theoretical insights from the literature on capital constrained lending (Kiyotaki and Moore 1997) are used to characterize firms’ borrowing constraints in a competitive, non-discriminatory equilibrium. The theory-based constraints are then transformed into an estimable stochastic frontier model under (arguably) weak assumptions about the borrower distribution. The null hypothesis of no distortion in credit availability can then be tested against the alternative hypothesis of “favoritism of state-owned firms” based on standard statistical techniques. For the purposes of this paper, the test differentiates between state-owned firms in general, and firms owned by local governments.

Following the literature, we formulate the borrowing constraint faced by firms as follows:
formula
1
where i indexes firms and t time, Loans is the total loan stock (in logs), Equity is equity capital (in logs) (which we use as a proxy for corporate net wealth), Distort is a vector of distortion indicators to be specified, is an independent normally distributed random variable, and the are the parameters to be estimated. The parameter may vary in time with general economic conditions, such as prices and interest rates, but it is constant across firms. We refer to the right-hand side of the borrowing constraint as the credit limit. Credit availability is the value of the limit.

In a competitive equilibrium without favoritism of state-owned firms we expect and , with indicating that credit availability is proportional to equity capital. Unit elasticity arises in some theoretical models but, in general, we interpret that theory does not provide clear guidance about the magnitude of The same holds for the sign and magnitude of the proportionality factor which absorbs all non-firm-specific variables that are not explicitly controlled. These include changes in the general price level, interest rates, and the invertors’ return requirement.

We include in Distort a state ownership indicator (gvt) to quantify favoritism of state-owned firms in general (i.e., both central and local government-owned firms). We add a separate indicator of local government ownership (lgvt): the parameter estimate of lgvt indicates whether credit availability of local government firms is different from that of state-owned firms in general. Under the null hypothesis of no favoritism of state-owned firms, the parameters for gvt and lgvt are both zero, which we test with the standard z test.

We also include indicators of sector of operation (Sector) and area (Area) in Distort. By including such indicators in the empirical model, we control for the effect of the government's objectives regarding the sectoral and geographical development of the Chinese economy on credit availability.

The borrowing constraint (1) is transformed into a stochastic frontier form without loss of generality. Define as the (inverse) distance of the loan stock from a credit limit:
formula
2
The borrowing constraint can then be written:
formula
3

We estimate equation (3) by maximum likelihood under the standard assumptions that v and u are independent random variables; v is normal; and u is either half normal, exponential, or truncated normal. The data are treated as a series of cross sections rather than a panel because panel techniques impose restrictions on the residual distributions that are difficult to rationalize from theory.

Rather than add more variables in the frontier vector as a “robustness check,” we rely on the theory-based parsimonious specification of the borrowing constraint (3) quite deliberately. The reason is that ad hoc variables, rather than functioning as “robustness checks,” may bias the results by hiding indirect discrimination at the loan market. A case in point is profitability, which, based on common sense, may significantly influence credit availability. This influence is not independent of favoritism of state-owned firms, however. Under the null hypothesis of no favoritism, there is no reason, for example, why current profitability should deviate systematically between private and state-owned firms. If, on the other hand, favoritism of state-owned firms prevails, then this likely has significant positive effects on their profitability. If profitability were included as an independent factor in the frontier vector, then it would capture this indirect effect of favoritism, and the test results with the state-ownership dummies will accordingly be downwards biased. For this reason, we adhere to the theoretical null of “no discrimination” and do not add ad hoc variables in the model.

To summarize, the econometric analysis is based on three main assumptions: the functional form of the limit, the distribution of v, and the distribution of u. Under these assumptions, a stochastic frontier analysis of the firm sample yields quantitative estimates of credit availability in parametric form. This allows us to quantify and test favoritism in credit availability toward state-owned firms in China.

4.  The data

The estimations are based on published balance sheet data of Chinese-listed companies. The focus on this sample of firms is justified by severe data quality issues, which characterize corporate accounting in China. For listed firms, the underlying data compilation methods are of relatively high quality because of listing requirements. A major modernization effort to harmonize the accounting standards of Chinese-listed companies with international norms was implemented in 1998. Although listed companies are a select group, they play a prominent role in the development of the Chinese economy. Government policies regarding macroeconomic developments in China are therefore likely reflected in credit availability of listed firms.

Table 1 summarizes the estimation data. The data provider is Wind Information Co., which specializes in data service provision for banks and investors operating in the local markets. The data covers all A-share companies (i.e., companies from mainland China) listed on the two mainland Chinese stock exchanges Shanghai and Shenzhen. Financial companies are excluded from the analysis. The number of companies included in estimations ranges between 1,000 and 2,000, depending on the year. The estimations are based on nominal series becasue changes in the general price level from year to year are captured in the estimations by the time varying regression constants. As our loan stock indicator, we use ‘financial debt,’ which includes bank loans and bonds. For corporate net wealth, we use ‘owners’ equity.’

Table 1. 
Descriptive statistics
yearDebt (%)Equity (%)east (%)snd (%)trd (%)gvt (%)cgvt (%)lgvt (%)
2004 11.07 11.13 63 70 26 88 85 
2005 11.17 11.14 64 70 26 87 84 
2006 11.16 11.10 65 70 26 84 80 
2007 11.09 11.11 66 71 25 78 75 
2008 11.20 11.28 66 71 25 78 75 
2009 11.35 11.51 66 71 25 78 75 
2010 11.49 11.87 66 71 25 78 75 
2011 11.66 11.97 66 71 25 78 75 
yearDebt (%)Equity (%)east (%)snd (%)trd (%)gvt (%)cgvt (%)lgvt (%)
2004 11.07 11.13 63 70 26 88 85 
2005 11.17 11.14 64 70 26 87 84 
2006 11.16 11.10 65 70 26 84 80 
2007 11.09 11.11 66 71 25 78 75 
2008 11.20 11.28 66 71 25 78 75 
2009 11.35 11.51 66 71 25 78 75 
2010 11.49 11.87 66 71 25 78 75 
2011 11.66 11.97 66 71 25 78 75 

Source: Wind Information Co. and authors' calculations.

Note: Sample means. Debt = financial debt (RMB 10000, log); Equity = equity capital (RMB 10000, log), east = eastern provinces; snd = second industry; trd = third industry; gvt = government ownership; cgvt = central government ownership; lgvt = local government ownership. Government ownership identified by ownership share.

Following the classification by the China Banking Regulatory Commission, state-owned firms (gvt) are identified by the type of the largest stockholder, also taking into account indirect ownership. This variable indicates state-owned firms in general, that is, both local and central government–owned firms. Local government firms (lgvt) are identified by diluting from gvt those firms that are owned by the central government in accordance with the State Owned Assets Supervision and Administration Commission of the State Council. State ownership share of all listed companies varies between 88 percent and 78 percent in the sample. Of this, the lion's share is by local governments. The remaining central government–owned firms cover about 4–5 percent of the firm sample.

The companies are classified by province of registration. Initially we aggregated the provincial level indicator into three areas, ‘east,’ ‘central,’ and ‘west’ in the standard way. About two-thirds of the firms originated in the eastern provinces, with the remainder about evenly split between western and central China. In the empirical analysis, we ended up aggregating the western and central areas for compactness without effect on the results.3

We use the three-level sector indicator: ‘Primary,’ ‘Secondary,’ and ‘Tertiary’ industry, based on the Classification of the National Economy Industries by the national statistical authority. In our sample, the Primary industry comprises mostly energy and mining firms, and accounts for only 4 percent of the sample. By far the largest group is the Secondary industry (70 percent of the total), which includes traditional manufacturing and electronics firms. The Tertiary industry includes service firms.

5.  Estimation results

Based on model stability and likelihood, we use the normal/half normal model as a benchmark. The first priority is to validate the approach by the skew criterion. The test results in Table 2, based on ten models estimated from cross sections of years 2004–11, provide strong evidence of residual skew for most years. The likelihood-ratio test rejects the null hypothesis of no skew at 5 percent for all years. Because the model includes a lagged term, year 2003 is excluded from the estimations.

Table 2. 
The benchmark models of credit limits
VARIABLES20042005200620072008200920102011
Equity, lagged 0.784*** 0.778*** 0.816*** 0.846*** 0.893*** 0.989*** 1.016*** 1.020*** 
 [0.0199] [0.0192] [0.0196] [0.0174] [0.0163] [0.0172] [0.0195] [0.0221] 
gvt 0.280*** 0.456*** 0.505*** 0.558*** 0.590*** 0.494*** 0.664*** 0.971*** 
 [0.106] [0.107] [0.110] [0.105] [0.101] [0.0984] [0.107] [0.112] 
lgvt −0.259*** −0.308*** −0.329*** −0.374*** −0.383*** −0.203** −0.0626 −0.0453 
 [0.0916] [0.0945] [0.0972] [0.0943] [0.0929] [0.0903] [0.0993] [0.105] 
east 0.0553 0.0757 0.0501 0.0464 −0.0460 −0.0769* −0.134*** −0.221*** 
 [0.0476] [0.0493] [0.0505] [0.0487] [0.0466] [0.0465] [0.0515] [0.0561] 
snd 0.145 0.142 0.308** 0.268** 0.249** 0.336*** 0.331*** 0.351*** 
 [0.116] [0.118] [0.121] [0.119] [0.112] [0.110] [0.122] [0.133] 
trd 0.0509 0.0484 0.113 0.0969 0.0683 0.200* 0.250* 0.331** 
 [0.121] [0.124] [0.127] [0.124] [0.117] [0.116] [0.129] [0.139] 
Constant 3.071*** 3.018*** 2.425*** 2.133*** 1.565*** 0.448* −0.0871 −0.325 
 [0.265] [0.259] [0.276] [0.243] [0.224] [0.231] [0.275] [0.318] 
Observations 1,285 1,373 1,403 1,546 1,789 1,793 1,788 1,789 
sigma_v 0.579 0.661 0.763 0.770 0.784 0.739 0.814 0.822 
sigma_u 1.005 0.962 0.776 0.779 0.797 0.920 1.021 1.269 
VARIABLES20042005200620072008200920102011
Equity, lagged 0.784*** 0.778*** 0.816*** 0.846*** 0.893*** 0.989*** 1.016*** 1.020*** 
 [0.0199] [0.0192] [0.0196] [0.0174] [0.0163] [0.0172] [0.0195] [0.0221] 
gvt 0.280*** 0.456*** 0.505*** 0.558*** 0.590*** 0.494*** 0.664*** 0.971*** 
 [0.106] [0.107] [0.110] [0.105] [0.101] [0.0984] [0.107] [0.112] 
lgvt −0.259*** −0.308*** −0.329*** −0.374*** −0.383*** −0.203** −0.0626 −0.0453 
 [0.0916] [0.0945] [0.0972] [0.0943] [0.0929] [0.0903] [0.0993] [0.105] 
east 0.0553 0.0757 0.0501 0.0464 −0.0460 −0.0769* −0.134*** −0.221*** 
 [0.0476] [0.0493] [0.0505] [0.0487] [0.0466] [0.0465] [0.0515] [0.0561] 
snd 0.145 0.142 0.308** 0.268** 0.249** 0.336*** 0.331*** 0.351*** 
 [0.116] [0.118] [0.121] [0.119] [0.112] [0.110] [0.122] [0.133] 
trd 0.0509 0.0484 0.113 0.0969 0.0683 0.200* 0.250* 0.331** 
 [0.121] [0.124] [0.127] [0.124] [0.117] [0.116] [0.129] [0.139] 
Constant 3.071*** 3.018*** 2.425*** 2.133*** 1.565*** 0.448* −0.0871 −0.325 
 [0.265] [0.259] [0.276] [0.243] [0.224] [0.231] [0.275] [0.318] 
Observations 1,285 1,373 1,403 1,546 1,789 1,793 1,788 1,789 
sigma_v 0.579 0.661 0.763 0.770 0.784 0.739 0.814 0.822 
sigma_u 1.005 0.962 0.776 0.779 0.797 0.920 1.021 1.269 

Source: Authors' calculations.

Note: The endogenous, left-hand side variable is Debt. The models are estimated separately for each cross section under the assumption that u is half normal. *Statistically significant at the 10 percent level; **statistically significant at the 5 percent level; ***statistically significant at the 1 percent level. sigma_v and sigma_u are the standard errors of v and u. p_z is the p-value of the likelihood-ratio test of H0: no residual skew. Standard error shown in parentheses.

Based on the skew criterion, we conclude in favor of the stochastic frontier model of borrowing constraints (3) for all years. The parameter estimates suggest variegated development of credit limits (Table 2). One trend is the increase in the marginal effect of equity from +0.8 in 2008 to +1.0 in 2011, accompanied by a decrease in the regression constant to insignificant levels. These changes suggest increased attention to borrowers’ solidity, possibly driven by improvements in risk assessment policies in Chinese banks. This interpretation accords with expert assessments, for example by the IMF, pointing to improved risk management practices among Chinese banks in line with international standards.

The estimations also signal a shift in the effect of geography on credit availability that accords with the government's objective to promote a more geographically balanced development of the Chinese economy. Earlier on in the period, companies in the relatively affluent eastern China, where the export industries are concentrated, had better credit availability than companies elsewhere. By 2011, the fixed effect of the eastern province had turned significantly negative, indicating that credit availability in eastern China had fallen below the rest of the country.

We observe that the fixed effects of Secondary and Tertiary industries turned significantly positive during the latter part of the sample. This implies that credit availability in these industries has improved relative to that of the Primary industry. Firms in the Primary industry are in our sample mostly energy and mining firms. The tightening of their credit conditions relative to firms in manufacturing and services is consistent with the government's environmental objectives, as well as the aim to expand the services (Tertiary) industry.

From the perspective of the ongoing debate about the Chinese model, an interesting finding is the increase in the fixed effect for state-owned firms, which signals an improvement in credit availability of state-owned firms relative to other firms. This “favoritism parameter” increased from about 0.3 at the start of the sample in 2004 to 0.97 in 2011. This estimation result is in line with the previous literature, and contributes to it by showing a strengthening of favoritism toward state-owned firms all the way up to 2011.

From the point of view of our research focus, an interesting result is the significant change in the fixed effect of local government owned firms. At the start of the sample in 2004, the local government fixed effect was significantly negative at −0.26, implying that credit availability of local government owned firms was significantly below that of state-owned firms in general, and roughly at par with private firms. The local government fixed effect shows an increasing trend throughout the sample period, and in 2011 it is statistically insignificant at −0.05. The increase of the local government ownership indicator from negative territory towards zero implies a large improvement in the credit availability of local government owned firms relative to other firms.

To summarize, the estimations quantify the changes in credit availability in time, across the main sectors, geographical areas, and ownership types in China. We observe growing favoritism of state-owned firms, and also a significant improvement in credit availability of local government owned firms relative to other firms. The results are consistent with the prevailing view that the policies of Hu Jintao's government increasingly favored SOEs. The estimation results are also consistent with a growing economic role of local governments in China. The results are robust to alternative distributions of u, omission of geographic and area indicators, heteroskedasticity, variable effects of equity across borrower groups, equity as a contemporaneous variable, nonlinear effects of equity, and omission of outliers based on extreme values of the debt to equity distribution.

6.  Discussion and concluding remarks

We study credit availability in China to reveal the underlying political trends in the Chinese economy. Building on novel estimation techniques and balance sheet data of listed firms during the years 2003–11, the estimations reveal a continuing push toward state capitalism and a further strengthening of the economic role of local governments. The estimations confirm the earlier findings in the literature about favoritism of SOEs in China. They contribute quantitative insight about the bias in credit availability in time, as well as novel insights about the changing role of local governments.

It should be emphasized that the evidence of favoritism of state-owned firms in credit availability does not implicate any specific channel of influence from political decision makers to bankers. In particular, the results do not imply that preferential treatment of state-owned firms is implemented explicitly by directive from government bodies to the bank directors. The Chinese central bank does give instruction to bankers about credit policy (via the so called “window guidance system”), but because the Chinese central bank is in favor of financial reform, and its instruction is typically general in nature, the window guidance system is unlikely to be the most important culprit behind increased preferential treatment of state-owned firms in China.

Based on anecdotal evidence, the channels of political influence on the banking system are likely to be more subtle. One such channel may be cooperation between public officials and bankers to promote overlapping personal interests. Political decision makers depend on bankers to finance the projects needed to maintain the growth momentum in their area of responsibility, and thereby promote their personal advancement in the party ranks. Bankers’ career prospects are in turn influenced by the goodwill of the political elite. In this setting it seems likely that the two reach a mutual understanding about credit arrangements, although how this understanding manifests in direct action in uncertain. Another channel of political influence on banks may be simply that banks, observing the general direction of political winds in China, bet on the likely winners.

Because favoritist policies have been in place for quite some time, it is likely that they have had significant and long-lasting effects on economic developments in China. The allocation of financial flows based on political rather than economic motives is likely to promote suboptimal investment decisions, and thereby eventually weigh on the potential of the economy to grow. The negative long-term effects of political intervention on economic efficiency have been confirmed in many studies. Regarding the Chinese case, Shiu (2002) estimates that the productive efficiency of SOEs in 1995 fell behind that of private firms by 4 to 16 percent, depending on the industry and location of operation. It is also well known that the potential for improvement in productivity of industrial firms in China is still substantial. Hsieh and Klenow (2009) estimate that the total factor productivity of Chinese firms between 1998 and 2005 dragged behind corresponding firms in the United States by 30–50 percent, in part due to the misallocation of financial resources. Previous research therefore indicates that the cost of favoritism of state-owned firms in China can be substantial.

The way political intervention influences economic behavior in the short to medium term is arguably strongly dependent on the objective of the intervention. In the Chinese case, the overriding economic objective of policymakers during the Hu era was the promotion of economic growth. This objective received significant emphasis in the five-year plans for the central government, and its effective implementation tended to rest with provincial and local-level government officials, whose career prospects depended on their ability to deliver the desired economic outcomes.

It therefore seems plausible that political influence on banks’ credit policies contributed to an increase in economic growth in the short run above its long-run potential, that is, an overheating of the Chinese economy. One may speculate that, to promote economic growth, Party officials may have colluded with bankers to obtain financing for state-owned firms. Agreement between the two opened the doors to “quasi fiscal stimulus,” an increase in investment by state-owned firms, building on ample financing at favorable terms. This proposition would explain the sharp increase in the investment to GDP ratio in China at the start of Hu Jintao's term in government.There is also anecdotal evidence that the companies owned by local governments have raised significant amounts of debt to support local development projects. The hypothesis is also consistent with the macroeconomic record: GDP growth in China during Hu Jintao's government exceeded official targets, and remained resilient even during episodes of significant international economic headwinds.

Although this quasi fiscal stimulus may have supported GDP growth in the short term, such policies are not sustainable in the long run. The inevitable termination of the quasi fiscal stimulus will lower investment demand in the long run and thereby push downwards both GDP growth and the investment to GDP ratio in China.

A key issue of future interest is to identify and quantify the long-run effects of favoritist policies. Future studies will, we hope, also shed light on the policies of the present government in this regard. Although the economic program of Xi Jinping's government emphasizes the market mechanism as a driver for economic development, it remains to be seen to what extent the principles are implemented in practice.

References

Bajona
,
Claustre
, and
Tianshu
Chu
.
2010
.
Reforming State Owned Enterprises in China: Effects of WTO Accession
.
Review of Economic Dynamics
13
:
800
823
.
Ding
,
Sai
,
Alessandra
Guariglia
, and
John
Knight
.
2013
.
Investment and Financing Constraints in China: Does Working Capital Management Make a Difference
?
Journal of Banking and Finance
37
(
5
):
1490
1507
.
Fungácová
,
Zuzana
,
Risto
Herrala
, and
Laurent
Weill
.
2013
.
The Influence of Bank Ownership on Credit Supply: Evidence from the Recent Financial Crisis
.
Emerging Markets Review
15
:
136
147
.
Guariglia
,
Alessandra
,
Xiaoxuan
Liu
, and
Lina
Song
.
2011
.
Internal Finance and Growth: Microeconomic Evidence of Chinese Firms
.
Journal of Development Economics
96
:
79
94
.
Héricourt
,
Jérome
, and
Sandra
Poncet
.
2009
.
FDI and Credit Constraints: Firm Level Evidence in China
.
Economic Systems
33
:
1
21
.
Herrala
,
Risto
.
2014
.
Forward-Looking Reaction to Financial Regulation
. European Central Bank Working Paper No. 1645.
Available at
http://www.ecb.int.
Hsieh
,
Chang-Tai
, and
Peter J.
Klenow
.
2009
.
Misallocation and Manufacturing TFP in China and India
.
The Quarterly Journal of Economics
CXXIV
(
4
):
1403
1448
.
International Monetary Fund (IMF)
.
2012
.
People's Republic of China: Staff Report for the 2012 Article IV Consultation
.
Available at
http://www.imf.org.
Kiyotaki
,
Nobuhiro
, and
John
Moore
.
1997
. Credit Cycles.
The Journal of Political Economy
105
(
2
):
211
248
.
Kowalski
,
Przemyslaw
,
Max
Büge
,
Monika
Sztajerowska
, and
Matias
Egeland
.
2013
.
State-Owned Enterprises: Trade Effects and Policy Implications
. OECD Trade Policy Papers No. 147.
OECD Economic Surveys
.
2013
.
China. Paris
:
OECD Publishing
.
Poncet
,
Sandra
,
Walter
Steingress
, and
Hylke
Vandenbussche
.
2010
.
Financial Constraints in China: Firm Level Evidence
.
China Economic Review
21
:
411
422
.
Shiu
,
Alice
.
2002
.
Efficiency of Chinese Enterprises
.
Journal of Productivity Analysis
18
:
255
267
.
Walter
,
Carl E.
, and
Fraser J. T.
Howie.
2011
.
Red Capitalism.
Singapore
:
John Wiley and Sons (Asia)
.
Zhao
,
Shiyong
.
2009
.
Government Policies and Private Enterprise Development in China: 2003–2006
.
China & World Economy
17
(
4
):
36
52
.

Notes

*

We thank the session participants in the CES Annual conference in Kaifeng, The “China Goes Global” conference in Harvard, the CEA conference in London, the Bofit seminar in Helsinki (all in 2012); and the Western Economic Society Pacific Rim conference in Tokyo and the Asian Economic Panel Meeting in Dublin (all in 2013).

1 

By state-owned firms, we mean firms that are majority owned either directly or indirectly (via investment companies) by the state.

2 

The methodology is presented in Herrala (2014). It has been used in a related context to study the loanable funds of banks by Fungácová, Herrala, and Weill (2013).

3 

The eastern area includes Beijing, Fujian, Guangdong, Hainan, Hebei, Jiangsu, Liaoning, Shandong, Shanghai, Tianjin and Zhejiang. Hong Kong, Canton, and Macau are excluded.