تحرک سرمایه منطقه ای در چین: اصلاحات اقتصادی با یکپارچگی مالی محدود
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14531||2013||11 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of International Money and Finance, Volume 37, October 2013, Pages 493–503
This paper assesses the changes in the regional capital mobility in China during the period of economic reform in 1978–2008 by employing a panel time varying coefficient (TVP) model. This approach is much more suitable to model China's evolution in the regional capital mobility than a standard structural break model as China's reforms took place gradually and were often implemented over several stages. Using the TVP model, we find that (1) China's provincial capital mobility demonstrated a moderate improvement over the sample period, but worsened temporarily between 1994 and 1997. This is probably due to the government's effort to combat inflation which reduced the investment and transfers to regions; (2) regions with the most developed and least developed provinces experienced higher degree of capital mobility improvement than those in the middle.
After Feldstein and Horioka (1980) measured international capital mobility by the correlation between national saving and investment, Feldstein, 1983, Murphy, 1984, Obstfeld, 1989 and Obstfeld, 1996, and Tesar (1991) documented relatively high saving and investment correlation among the OECD countries. The estimates were generally above 0.5, and in some cases statistically insignificantly different from unity. This evidence for low international capital mobility ran counter to the wildly held consensus of relatively high international capital mobility. To reconcile this empirical result with the assumption of high international capital mobility, Cantor and Mark, 1988, Mendoza, 1991, Backus et al., 1992 and Baxter and Crucini, 1993 incorporated adjustment costs for domestic investment, as well as other frictions, into models which also assumed free capital mobility. Coakley et al. (1996) argued that these cross-country studies are subject to the current account solvency constraint, which complicates the interpretation of the saving retention rate. As such, another group of papers apply the Feldstein and Horioka (1980) framework to intra-national cases. For example, Bayoumi and Rose, 1993, Helliwell and McKitrick, 1999 and Yamori, 1995, and Dekle (1996) use the framework to study provincial and regional capital mobility within a country, which generally is not subject to the same constraint. The first two studies focus on UK and Canada respectively, and the latter two on Japan. Although the Feldstein and Horioka (1980) framework has been widely adopted as a measure of capital mobility, it falls short in its theoretical foundation. By adopting the permanent income model, with a fraction of current income consumers, originally developed by Campbell and Mankiw, 1990a, Campbell and Mankiw, 1990b and Shibata and Shintani, 1998 constructed an alternative way to measure capital mobility. Instead of focusing on the correlation between national saving and investment, they estimate the correlation between national consumption and net output under the theoretical framework of a small open economy with permanent income consumers, and restrictions on capital mobility. Decressin and Disyatat (2008) extended this framework and measured capital market integration within the OECD and across Canadian provinces when liquidity-constraint agents are present. Huang (2010) further incorporated the terms-of-trade factor into the framework and estimated the degree of capital mobility controlling for the effect of the terms-of-trade. In general, the advantage of such a framework over the Feldstrein–Horioka approach is that it has a solid theoretical model embedded in the partial equilibrium permanent income setup. In this paper, we gauge the regional capital mobility across China by utilizing the above framework of Campbell and Mankiw, 1990a and Campbell and Mankiw, 1990b over the sample period of 1978–2008. Since China is now the world's second largest economy with a huge current account surplus against its major trading partners, a study of this topic is of importance in several aspects. First, China started its economic reform and opening up policy in 1978. During the last three decades, China has been experiencing a high rate of economic growth, up to 10% on average. Since this reform is market-oriented, many are eager to find out if any integrated capital markets across China has been established. Second, the growth miracle of China is driven largely by the fast growth in the export-oriented industries along the coastal line of China, while the vast west region is left out 1( Fleisher et al., 2010). Has this export performance produced any effect on the cross region capital mobility? Since China is now under intense international pressure to balance its huge current account surplus, export dependency will have to give way to enhanced domestic consumption demand in the future. For this reason, a more integrated capital market across regions would facilitate the effectiveness of the stimulus policies which aimed at promoting domestic demand and global rebalancing. Our results show that in the absence of widespread financial integration, a global stimulus package originating from China may not have any prominent effect, since changes in the domestic consumption across regions are closely tied to changes in the regional net income. Much effort has been devoted to evaluating the financial and capital market integration in China. Boyreau-Debray and Wei (2004) applied the Feldstein and Horioka (1980) framework to the provincial level saving and investment data in China from 1952 to 2001 and found that there was no improvement in the provincial capital mobility across China over this period. Using the same framework, Li (2010) applied the panel cointegration method to the provincial data from 1978 to 2006. He came to the same conclusion that capital market integration within China was low and there was practically no improvement during the sample period. Chan et al. (2011b), using the framework of both Feldstein and Horioka (1980) and Shibata and Shintani (1998), evaluated the provincial capital mobility in China during 1970–2006. They found that although the correlation between provincial savings and investment and that between provincial consumption and net output in China are high which signal strong barriers to capital mobility, these correlations declined after 1990s. This shows signs of improved capital mobility in recent period. Chan et al. (2011a), using a bootstrap panel cointegration analysis on savings–investment relationship that allows for cross-sectional dependence arising from nation-wide policies and shocks, also found that the degree of regional capital mobility across regions in China increased after 1997. Besides, researchers have also gauged the degree of financial market integration by implementing tests on the inter-provincial consumption risk sharing. They measure how effective the national financial market is in providing consumption insurance against idiosyncratic income risks to provinces. Xu (2008), by using provincial household survey data, found that consumption risk sharing across provinces in China is quite high, which indicates a rather high degree of financial integration within China. This result is supported by Feng (2010). In contrast, Curtis and Mark (2010) and Chan et al. (2012) found that the inter-provincial consumption risk sharing in China was low and was comparable to that across the OECD countries. Du et al. (2011) and Chan et al. (2012) further showed that there was a decline in the provincial risk sharing in China during the more recent period from late 1990s to 2008. We differentiate our study from previous ones in that we use a panel time-varying coefficients model to evaluate the degree of capital mobility across provinces and regions in China. This framework allows the crucial parameter that captures the degree of inter-provincial capital mobility to evolve over time. Since China has undergone a series of economic reforms in the last few decades, allowing time varying coefficients can more appropriately reflect the continuous process of structural change. Our results indicate that there was a slight improvement in capital market integration across different regions in China. Moreover, for more developed regions like the Capital Region (which includes Beijing, Tianjin and Hebei) and the Yangzi River Delta Region (which includes Shanghai, Jiangsu and Zhejiang), our estimation results indicate that they have a low degree of capital market integration with the rest of the country. The paper is organized as follows: Section 2 presents the theoretical framework. Section 3 discusses the dataset and the results from a benchmark constant coefficient panel data model. Section 4 presents the time-varying coefficient model and contrasts the estimation results with the benchmark model. Section 5 concludes.
نتیجه گیری انگلیسی
This paper examines the degree of capital mobility across different regions in China during 1978–2008 by utilizing the permanent income model developed by Campbell and Mankiw, 1990a and Campbell and Mankiw, 1990b. What is new in this paper is the use of an econometric technique that allows for different ”saving-retention” coefficients over time and across regions. Given that much has changed in China over time, this is a useful tool for assessing the process of financial integration in China. In our analysis, we abstracted from developments in economic geography and the related agglomeration literature in regional economics. In particular, Herrerias and Ordoñez (2012) have reported that provinces of different regions have converged into ”clubs”, based on measures of per-capital output, labor productivity and capital intensity. Their study notes considerable divergence in the underlying fundamentals for balanced economic growth in China. Their findings are consistent with our empirical results that document different degrees of financial integration across regions and time. The constant λ model yields positive and insignificant estimates of λ for the whole sample period and the two subperiods. However, when we employ a panel time-varying coefficients model, the results show that there is a slight improvement in the capital market integration across different regions in China. However, the two most developed regions, Beijing, Tianjin and Hebei as well as Shanghai, Jiangsu and Zhejiang show relatively large λ estimates, indicating that there was still some segregation between these developed regions and the other regions in China.