Skip to main content

Advertisement

SpringerLink
Log in

The size of the SOE sector and macroeconomic performance: an empirical study based on Chinese provincial data

  • Published:
Economic Change and Restructuring Aims and scope Submit manuscript

Abstract

Using a neoclassical growth model augmented with human capital, we investigate the impact of the presence of state owned enterprises (SOEs) on macroeconomic performance in China, using provincial data from 1990 to 2004. We estimate a macroeconomic model with panel methods to explain changes in labor productivity resulting from standard influences as well as presence of the SOE sector measured in five different ways. While controlling for growth in the labor force and physical capital, government size, exposure to trade and change in economic structure, we conclude that the relative share of the SOE sector has no significant influence on macroeconomic performance in China during our sample period.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. It is worth noting that “state-owned units” includes the SOE sector as well as other governmental organizations. Therefore, as we show later, this investment share is not a proper measure of the relative size of the SOE sector. Tobin (2005, p. 734) distinguishes three classes of state-owned units: administrative agencies (which have power over other organizations), service organizations (which lack such power yet do not aim to make profit, such as hospitals and schools), and state enterprises. In our paper the SOE sector is refers to only this last type of unit.

  2. However, the authors did not describe the data sources they used. For example, it is not clear what “output value” is since there are two versions in official data sources, i.e. “gross output value” and “net output value.” Only the latter should be compared with provincial income. But there are problems in the availability of this indicator during their sample period.

  3. This indicator that Lin and Liu use to measure the SOE presence in the economy is problematic because the Chinese statistical authority introduced a new accounting method in 1998. We address this matter below in subsect. 3.1.

  4. As we show later, these two indicators are not proper measures of the relative size of the SOE sector.

  5. Note that we use y to represent labor productivity (Y/L) and \( \hat{y} \) to represent output per effective labor unit (Y/AL), following notation in Islam (1995).

  6. Given how we treat time in our regressions (see Sect. 3.1), τ = 4 years.

  7. Mankiw et al. (1992) and Islam (1995) provide extensive discussion of convergence as analyzed by models like this. Our regression results indicate that convergence was occurring across provinces.

  8. One exception to this statement is that we use government expenditures in some of our equations, and this is also included in the panel data evaluated in Narayan et al. (2008).

  9. Narayan 2008 finds that allowing for endogenously determined breaks in real GDP and per capita real GDP for panels on 24 provinces from 1952 to 2003 allows him to conclude panel stationarity for both the whole set of provinces as well as three regional blocks (eastern, central and western). Narayan et al. 2008 examine the same data without allowing for structural breaks. They find the panels for both real GDP and per capita real GDP contain unit roots in levels but not first differences.

  10. The research on stationarity cited in this paragraph used Chinese data series covering 47 and 52 years.

  11. These include our dependent variable, Y/L, the measure of SOE presence based on state share of labor, L S /L, and the variable we use to measure structural change, STR.

  12. These years are 1991, 1996 and 2001.

  13. Although SOEs are, in some sense, part of the government, they operate on a commercial basis whereas government agencies do not.

  14. We use the second rather than first year to avoid the problem discussed in Sect. 3.1 regarding employment data.

  15. Although there is a consensus among economists about the central role played by human capital in the process of economic growth, the data for measuring the stock of human capital or investment in human capital are far from satisfactory. Our framework avoids the need to measure the stock of human capital, but we have to measure the investment ratio of human capital, i.e., the portion of output devoted to human capital accumulation. Mankiw et al. (1992) used the proportion of middle school students in work-age population as an indicator of the investment ratio of human capital. They note that if such measures are proportional to the theoretical ideal, s H , then they can be employed as a proxy variable in regression equations based on aggregate production relations. Other authors followed this treatment using the same or similar frameworks (Islam 1995; Knowles and Ozanne 2003).

  16. In fact CSY does not directly report the value added data for 1998. We derived it using gross value of industrial output for state-owned and state-held industrial enterprises and the corresponding ratio of value added (i.e. the ratio of value added to gross output value).

  17. From 1998 forward data for “state-owned” units and enterprises is described as “state-owned and state-held enterprises.” The meaning of “state-held” is that the state owns a majority of voting shares. The essential point for both is that the state has legal administrative control.

  18. For a few variables in some sub-periods only four years of data are available, so their averages are for four years. This includes V S /V and V S /Y in the periods 1990–1994 and 2000–2004. Because STR is the change in structure (employment share of secondary and tertiary sectors) and employment data in census years is not usable, it is constructed as the change across the last 4 years of each 5-year time period. For example, for the sub-period of 2000–2004, STR is the change from 2001 to 2004. See the end of Sect. 3.1 for an explanation of the data problem in census years.

  19. The econometric software we used is Stata8.0. We calculated Hausman statistics and F-statistics for fixed effects and find that they mostly support the fixed-effect estimates. The exceptions are equations 2-1, 4-3, 4-5, 5-1, 5-7, 6-1 and 6-3 in which the Hausman statistics are negative, and equation 1-7, for which the statistic is not significant. However, with the exception of equation 6-3, the Hausman statistics for the corresponding unrestricted equations strongly support the fixed effect estimates, and for these seven equations, the F-statistics for the restriction are all small. This implies that the restriction holds approximately and so the fixed effect estimates are reliable for the equations with negative or insignificant Hausman statistics.

  20. It should be noted that we report and discuss only one-way fixed effect estimates, that is, models without time dummies. In fact, we have carried out two-way estimations for each equation. There are some differences between these two kinds of estimations. For example, the implied convergence speeds are faster. However, for what we are most concerned about, the coefficients of variables representing the relative size of the SOE sector, there is no essential difference. To save space we do not report and discuss these estimates.

  21. Islam (1995) and Dawson (1998) also find that human capital lacks statistical significance using a similar framework. In a study of conditional convergence across China’s regions Dong (2004) finds that the percentage of school-age children enrolled is not significant. In other studies the effect of human capital measured by years of education is disputable. See, for example, Bils and Klenow (2000). Mankiw et al. (1992) found that investment in human capital had statistical significance in their cross-country regressions (which did not include China). As Islam discusses (1995, p. 1153) when models “…incorporate a temporal dimension of human capital …” the variables typically perform poorly in a statistical sense.

  22. Two recent papers examine the relevance of Wagner’s Law in China’s development since 1978. Narayan et al. (2008) use provincial data in a simple model and find weak confirmation that Wagner’s Law applies in only the western region. Tobin (2005) engages in a more involved analysis of the applicability of the Law to China and finds that Chinese development exhibits the same patterns that Wagner observed in Europe in the nineteenth century. However, Tobin does not explain the mechanism behind the positive correlation between national wealth and government expenditure, nor does he explicitly treat China’s distinction in having a relatively large SOE sector. Even today in China conventional roles of the state are often played out through SOEs in a disguised form.

  23. China Statistical Yearbook (2005), Table 5–9, p. 130.

  24. The accounting standard for value added from SOEs actually changed twice during our time frame. In 1992, it changed from “net output value of publically-owned industrial enterprises with independent accounting systems” to “value added of state-owned industrial enterprises with independent accounting system.” And in 1998, it changed again to “value added of state-owned and state-held industrial enterprises”.

  25. As mentioned above, the value added measures are not ideal indicators. Although the inconsistency of statistical standards is not a problem, the data exclude some SOEs (e.g. SOEs in the tertiary sector are left out). Therefore, they understate the size of the SOE sector. However, they are the best indicators available at present.

  26. Mankiw et al. (1992) and Islam (1995) found the same result regarding the restriction.

References

  • Aron J (2000) Growth and institutions: a review of the evidence. World Bank Res Obs 15(1):99–135

    Google Scholar 

  • Barro R (1997) Determinants of economic growth: a cross-country empirical study. The MIT Press, Cambridge, Massachusetts

    Google Scholar 

  • Bils M, Klenow PJ (2000) Does schooling cause growth? Am Econ Rev 90(5):1160–1183

    Article  Google Scholar 

  • Chen B, Feng Y (2000) Determinants of economic growth in China: private enterprise, education, and openness. China Econ Rev 11:1–15

    Article  Google Scholar 

  • Dawson JW (1998) Institutions, investment, and growth: new cross-country and panel data evidence. Econ Inq 36:603–619

    Article  Google Scholar 

  • Doamekpor KF (1998) Contributions of state-owned enterprises to the growth of total output. Int Econ J 12(4):65–77

    Article  Google Scholar 

  • Dong X (2004) An analysis of the regional income inequalities of China: 1952–2002. J Econ Res (Chinese). No. 9:48–59

  • Gylfason T, Herbertsson TT, Zoega G (2001) Ownership and growth. World Bank Econ Rev 15(3):431–449

    Article  Google Scholar 

  • Huang X, Ping L, Lotspeich R (2008) Efficiency and externality effects of state-owned enterprises on economic growth: an analytic framework and empirical study based on Chinese provincial data. Working Paper, Center for Comparative Economic System at Liaoning University

  • Islam N (1995) Growth empirics: a panel data approach. Q J Econ 110(4):1127–1170

    Article  Google Scholar 

  • Jalilian H, Weiss J (1997) Bureaucrats, business and economic growth. J Int Dev 9(6):877–885

    Article  Google Scholar 

  • Jefferson GH (1998) China’s state enterprises: public goods, externalities and Coase. Am Econ Rev—Papers and Proceedings 88(2):428–432

    Google Scholar 

  • Jefferson GH, Rawski TG, Li W, Yuxin Z (2000) Ownership, productivity change, and financial performance in Chinese industry. J Comp Econ 28:786–813

    Article  Google Scholar 

  • Knowles S, Garces-Ozanne A (2003) Government intervention and economic performance in East Asia. Econ Dev Cult Change 51(2):451–477

    Article  Google Scholar 

  • Levine R, Renelt D (1992) A sensitivity analysis of cross-country growth regression. Am Econ Rev 82(4):942–963

    Google Scholar 

  • Lin S (2000) Resource allocation and economic growth in China. Econ Inq 36(3):515–526

    Google Scholar 

  • Lin J, Liu M (2003) The convergence of economic growth and income distribution in China. World Econ (Chinese) No. 8:3–14

  • Liu X (2000) The effect of ownership structures of industrial enterprises of China on differences of efficiencies: an empirical analysis based on data of National Industrial Census in 1995. J Econ Res (Chinese) No. 2:17–25

  • Liu Y (2001) On the macroeconomic efficiencies of state owned enterprises. Soc Sci China (Chinese) No. 5:69–81

    Google Scholar 

  • Mankiw NG, Romer D, Weil DN (1992) A contribution to the empirics of economic growth. Q J Econ 107(2):407–437

    Article  Google Scholar 

  • Narayan PK (2004) Are output fluctuations transitory? New evidence from 24 Chinese provinces. Pac Econ Rev 9(4):327–336

    Article  Google Scholar 

  • Narayan PK (2008) Evidence of panel stationarity from Chinese provincial and regional income. China Econ Rev 19:274–286

    Article  Google Scholar 

  • Narayan PK, Nielsen I, Smyth R (2008) Panel data, cointegration, causality and Wagner’s law: empirical evidence from Chinese provinces. China Econ Rev 19:297–307

    Article  Google Scholar 

  • National Bureau of Statistics of China (2005) Comprehensive statistical data on 55 years of People’s Republic of China. China Statistical Publishing House, Beijing

    Google Scholar 

  • National Bureau of Statistics of China (various years) China statistical yearbook. Beijing: China Statistical Publishing House

  • Palia D, Phelps E (2000) The empirical importance of private ownership for economic growth. In: Paganetto L, Phelps E (eds) Finance, research, education and economic growth. Macmillan, London, pp 25–37

    Google Scholar 

  • Plane P (1992) Productive efficiency of public enterprises: A macroeconomic analysis based on cross-section estimation of a neoclassical production function. Appl Econ 24:822–844

    Article  Google Scholar 

  • Ram R (1986) Government size and economic growth: a new framework and some evidence from cross-section and time-series data. Am Econ Rev 76(1):191–203

    Google Scholar 

  • Sachs J, Woo W (1997) Understanding China’s economic performance. NBER Working Paper No. 5935

  • Smyth R, Inder B (2004) Is Chinese provincial real GDP per capita nonstationary? Evidence from multiple trend break unit root tests. China Econ Rev 15(1):1–24

    Article  Google Scholar 

  • Temple JRW (2005) Dual economy models: a primer for growth economists. Manchester Sch 73(4):435–478

    Article  Google Scholar 

  • Temple JRW, Woessmann L (2006) Dualism and cross-country growth regressions. J Econ Growth 11:187–228

    Article  Google Scholar 

  • Tian X (2001) Privatization and economic performance: evidence from Chinese provinces. Econ Syst 25:65–77

    Article  Google Scholar 

  • Tobin D (2005) Economic liberalization, the changing role of the state and Wagner’s Law: China’s development experience since 1978. World Dev 33:729–743

    Article  Google Scholar 

  • World Bank (1995) Bureaucrats in business: the economic and politics of government ownership. Oxford University Press, Oxford

    Google Scholar 

Download references

Acknowledgments

This research was supported by Ministry of Education of China (No. 05JJD790009) and Bureau of Education of Liaoning Province (No. J05068). We gratefully acknowledge the comments of Ding Lu and an anonymous referee on an earlier version of the paper as well as econometric advice from Dale Bremmer. Remaining errors are ours alone.

Author information

Authors and Affiliations

  1. School of Economics, Liaoning University, 110036, Shenyang, Liaoning, People’s Republic of China

    Xianfeng Huang

  2. Center for Comparative Economic Systems, Liaoning University, 110036, Shenyang, Liaoning, People’s Republic of China

    Ping Li

  3. Department of Economics, Indiana State University, Terre Haute, IN, 47809, USA

    Richard Lotspeich

Authors
  1. Xianfeng Huang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ping Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Richard Lotspeich
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Richard Lotspeich.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Huang, X., Li, P. & Lotspeich, R. The size of the SOE sector and macroeconomic performance: an empirical study based on Chinese provincial data. Econ Change Restruct 42, 319–343 (2009). https://doi.org/10.1007/s10644-009-9074-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10644-009-9074-8

Keywords

JEL Classification

Search

Navigation