پس انداز، سرمایه گذاری، و تحرک سرمایه در چین

This paper addresses the capital mobility among regions within China. Using a range of panel estimators which deals with the non-stationarity of time series components, individual heterogeneity and common unobserved factors, we show that the savings and investment (both expressed as ratios to GDP) are positively correlated for a sample of 28 Chinese provinces over the period of 1978 to 2006. According to the Feldstein-Horioka's argument (1980, Economic Journal (90), pp.314–329), such a correlation can be interpreted as evidence of low capital mobility. In addition, by means of Granger causality test, we fail to provide consistent evidence to support the hypothesis of efficient capital allocation in China. Combining the results given above, it is believed that the capital may be inefficiently retained within the provincial confines. We conjecture that the intermarriage between financial power and local authorities is primarily responsible for this worrying phenomenon.

To some extent, the opening of China's ‘inner doors’, namely domestic market integration, as an integral part of the post-1978 economic reform, had long been neglected and overshadowed by the relatively successful opening up to the outside. In particular, it is argued that the decentralization of economic management and decision-making undertaken over the 1980s considerably weakened China's central government's capacity to control regional affairs and facilitated the local protectionist practices (Wang and Hu, 2001 and World Bank, 1994). However, partly due to the 1989 political crisis and the collapse of Soviet Union in 1991, which posed a serious challenge to the unity of China, the issue of national market integration has received increasing attention from both policymakers and economic scholars since the early 1990s. On the one hand, the Chinese authorities, especially the central government, have started their determined efforts to break down localism and set up a unified domestic market, including the center-region tax assignment reform, financial institutions restructuring and the vertical administration (chuizi guanli) reconstitution (see Yang (2004) for a detailed discussion). On the other hand, a vast body of empirical literature has emerged. Without reaching a consensus on the extent or even the trends on interregional integration over reform period, many facets of domestic market development have been explored, such as price convergence ( Fan and Wei, 2006, World Bank, 1994 and Young, 2000), production structure similarity ( Bai et al., 2004, World Bank, 1994 and Young, 2000), synchronization of the business cycle ( Poncet and Barthélémy, 2008, Tang, 1998 and Xu, 2002), internal trade links ( Naughton, 2003, Poncet, 2005 and World Bank, 1994), and labor migration ( Liang and Ma, 2004, Poncet, 2006 and Wei, 1997). 1 Nevertheless, to the best of our knowledge, few studies have focused on the integration of China's capital market, with the notable exception of Boyreau-Debray and Wei (2004, hereafter BDW) which address the capital mobility across regions. Using a panel of Chinese provinces (or province-level regions) for the period of 1978–2001, they find that provincial savings and investment rates are closely linked. According to the seminal work of Feldstein-Horioka (1980, hereafter FH), these results could be interpreted as evidence for low degree of interprovincial capital mobility. Because of the relative scarcity of capital stock in China, the frictionless allocation of such limited financial resources toward more productive areas appears to be of paramount importance. Roughly speaking, there are at least four reasons worth noting. First, Chinese enterprises, especially the non-state-owned ones, require a liberalized capital market where all kinds of origin-related discrimination are absent. Such a market can not only ease external funds constraints facing them, but also facilitate their expansion in respond to the nationwide opportunities. Second, to seek better return-risk trade-off, investors also need to freely diversify their financial assets across regions which may suffer from heterogeneous shocks. Third, by the similar logic of the goods market integration, additional gains would be associated with financial institutions reorganization: the least efficient ones should be eliminated from a highly integrated market, whereas the remaining ones will achieve economies of scale. Fourth, as the nascent consumption credit market is emerging in China, high capital mobility will equally matter for the consumption smoothing. Focusing on the issue of capital mobility, we attempt to reexamine the Chinese evidence over the past three decades. Compared to the study of BDW (2004), the current paper has three major distinguishing features. First, using an updated sample over 1978–2006, we show that, contrary to the results given in BDW (2004), both provincial savings and investment rates appear to be individually non-stationary and there is no strong evidence for the co-integrating relation between them. It is argued that such findings have crucial implications for the long run savings-investment association. Second, to deal with the regional heterogeneity and national common factors, we implement a set of panel estimators, which differs from those used in BDW (2004). More importantly, our estimates remain consistent under the time series properties mentioned above. Third, as will be detailed later, the investigation on capital allocation efficiency provided in BDW (2004) relies on a problematic logic that in the case of efficient market, the capital flows are allocated toward regions with higher marginal product of capital. Alternatively, we propose examining the causal relationship between aggregate investment and income growth rates in the hope of shedding light on the allocative efficiency of China's capital market. The reminder of this paper is organized as follows: Section 2 briefly presents the FH approach as well as some alternative interpretations on it. Section 3 examines the time series properties of China's savings and investment rates. Section 4 shows the results of FH regressions from a variety of panel estimators. A short discussion of the findings in other FH studies is also provided. Section 5 investigates further the efficiency of capital allocation in China. The last section concludes the paper.

This paper addresses the interregional capital mobility within China, which can be viewed as a major indicator of financial market integration. In the spirit of FH (1980)'s study, we first assess the degree of capital mobility during the reform era by examining the correlation between savings and investment rates. Using various panel estimators that tackle the non-stationarity of time series components, province-specific heterogeneity and unobserved common factors, we find a significant positive savings-investment association for a sample of Chinese provinces over the period of 1978 to 2006. According to the view of FH, these results can be interpreted as evidence of low degree of capital mobility. In other words, a part of incremental savings is always retained inside a province to fuel local investment. Moreover, by means of Granger causality test, we next investigate the efficiency of capital allocation in China. Focusing on the bilateral causation relationship between aggregate investment growth and income growth, we fail to provide consistent evidence to support the hypothesis of efficient capital allocation for most provinces. So, it seems reasonable to conjecture that the capital cannot go anywhere in search of the best return-risk trade-off and hence, is inefficiently constrained within the provincial confines. Finally, one may ask why the degree of interregional capital mobility is low within China. A natural explanation for that consists of the intermarriage between financial power and local authorities. Until the late 1990s, not only the provincial and municipal branches of the central bank, the People's Bank of China, but also those of commercial banks were subject to persistent intervention and even dual leadership of the local governments. As Yang (2004) discusses, the local officials seeking high growth rates (on which they tend to be evaluated) eagerly exerted pressures on state banks to lend to local enterprises, irrespective of economic fundamentals. In the extreme, Qian (2000) argues that the local government even had the authority to decide whether an enterprise should pay back a loan in the 1980s. Consequently, a rising and large share of bank loans are nonperforming and the major state banks are technically insolvent (Lardy, 1998). Given the fact that the banks dominate the financial sector in China, the distortion in loans remains serious obstacles to efficient capital mobility across regions. Furthermore, the development of the market-based direct financing channels, such as securities markets that appear to be more transparent and less plagued by the problems of local protectionism than bank lending, is still in its initial stage. Until recently, equities and bonds play a minor role in mobilizing and pooling financial resources. From the data of the People's Bank of China10, until 2005, they represented only 14.1% of total sources (stock) of the non-financial corporations external financing, while bank loans accounted for 64%. At the same time, households invested only 8.6% of their financial assets in the form of securities, while 72% in deposits. Nevertheless, since the late 1990s, more radical financial reforms, including the reorganization of the central bank, the centralization of the securities markets regulation and the recent joint-stock reform for the state commercial banks, have been launched in the hope of enhancing the role of market forces in capital allocation. Clearly, despite all these welcome efforts, China's economic revolution, as argued by Lardy (1998), is far from finished.