Efficiency of Chinese Township and Village Enterprises and Property Rights in the 1990s: Case Study of Wuxi

Go Yano; Maho Shiraishi

doi:10.1057/palgrave.ces.8100037

Comparative Economic Studies

June 2004, Volume 46, Issue 2, pp 311–340 | Cite as

Efficiency of Chinese Township and Village Enterprises and Property Rights in the 1990s: Case Study of Wuxi

Authors
Authors and affiliations

Go Yano
Maho Shiraishi

Regular Article

First Online: 09 June 2004

5 Downloads
4 Citations

Abstract

We estimate the technical efficiency of collective-owned township and village enterprises (TVEs) relative to other types of enterprise, taking into account the net work rate for equipment and machines. Our aim is to investigate whether the vaguely specified property rights of these enterprises caused them to decline in technical efficiency just before they underwent a massive privatisation programme in Wuxi City and Jiangsu province, starting in 1998. Our results cast doubt on the view that vaguely specified property rights became inconsistent with productive or technical efficiency in Wuxi City. It follows that privatisation of collective-owned TVEs may be insufficient to improve their productivity performance from now on.

Keywords

collective-owned TVE property rights technical efficiency

JEL Classifications

O53 P21 P32

This is a preview of subscription content, to check access.

Appendix A: Estimation results by production functions with firm-specific fixed effects

As stated in footnote 9, in order to find out whether our estimates of coefficients are biased, we now re-estimate the production function using the panel estimation method. If the resulting estimates are consistent with those in the text, the latter are unbiased.

Here, we estimate the production functions with firm-specific fixed effects using the same micro data as in the test:

Open image in new window

and

Open image in new window

where c_i is the firm-specific fixed effect of firm i and the other notation is as in the text. A random effects model cannot be adopted since in every case the Hausman–Wu test rejects the null hypothesis that the model specification of production function with firm-specific random effects is valid.

The estimation procedure for the production functions (A.1–A.3) is the same as in the text: simultaneous estimation of each production function with equation (4) in the text, and the regression equations of ln K and ln L to instrumental variables by the 3SLS method, taking heteroscedasticity into account.

The estimates of production functions (A.1–A.3) are presented in Table A1. For ease of comparison, estimates of the production function without NWR and with firm-specific fixed effects are also reported in Table A1.

Table 4

Estimates of production functions with firm-specific fixed effects^a^,^b^,^c

	Without NWR	Equation (5)	Equation (6)	Equation (7)
	Time dummy variables
1992	0.481^**	0.214^**	0.211^**	0.186^**
	(7.269)	(3.502)	(3.432)	(2.816)
1993	0.946^**	0.598^**	0.593^**	0.515^**
	(13.619)	(5.326)	(5.184)	(6.444)
1994	0.893^**	0.627^**	0.720^**	0.680^**
	(11.148)	(7.924)	(7.742)	(6.520)
1995	0.943^**	0.991^**	0.784^**	0.734^**
	(9.813)	(8.488)	(7.318)	(7.539)
1996	0.987^**	1.334^**	1.026^**	1.174^**
	(9.420)	(10.762)	(8.540)	(9.254)
1997	0.918^**	1.325^**	1.034^**	1.109^**
	(7.050)	(9.588)	(8.454)	(7.201)
ln K	0.721^**	0.831^**	0.833^**	0.807^**
	(8.963)	(23.063)	(22.822)	(20.831)
ln L	0.378^**	0.189^**	0.197^**	0.217^**
	(5.034)	(2.280)	(2.330)	(2.592)
ln(M/K)		0.385^**	0.436^**	0.160^**
		(10.368)	(7.825)	(2.736)
	ln(M/K) × Enterprise type dummy variables
ln(M/K) × SOE			−0.200^**	−0.128^*
			(−2.766)	(−2.196)
ln(M/K) × SHE			−0.153^*	−0.126^*
			(−2.543)	(−2.280)
ln(M/K) × PE			0.044	0.085
			(1.223)	(1.251)
ln(M/K) × JV			−0.071	−0.133
			(−1.891)	(−1.683)
ln(M/K) × UCOE			−0.062^*	−0.233^*
			(−1.989)	(−2.446)
ln(M/K) × UCE			−0.047^**	−0.019^*
			(−2.941)	(−2.205)
ln(M/K) × UOE			−0.303	−0.540
			(−1.532)	(−1.299)
	ln(M/K) × Time dummy variables
ln(M/K) × 1992				0.223^**
				(2.930)
ln(M/K) × 1993				0.256^**
				(3.118)
ln(M/K) × 1994				0.281^**
				(3.652)
ln(M/K) × 1995				0.363^**
				(4.838)
ln(M/K) × 1996				0.496^**
				(6.680)
ln(M/K) × 1997				0.537^**
				(6.927)
Hausman–Wu test^d	66.592	87.918	94.424	109.400
P-value	0.000	0.000	0.000	0.000
Adjusted R²	0.730	0.736	0.739	0.751
Adjusted GR^2e	0.683	0.691	0.692	0.703
Number of observations	2018	2018	2018	2018

^aEach of the production functions (5)–(7) is estimated simultaneously with equation (4) and the regression equations of ln K and ln L to the instrumental variables using the 3SLS method. The production function without NWR is estimated simultaneously with just the regression equations of ln K and ln L using the 3SLS method. The joint estimates of equation (4) and the regression equations of ln K and ln L are not reported.

^bThe table presents regression coefficients. The dependent variable is ln Y (log output). Numbers in parentheses are the t-statistics.

^cThe estimates of firm-specific fixed effects are omitted to save space.

^dThe Hausman–Wu test is the test value for the specification of random and fixed-effects models.

^eThe GR² is the measure of goodness-of-fit of the regression models estimated by the instrumental variables (IV) method, generalized R², proposed by Pesaran and Smith (1994) as better than ordinary R². Adjustment of the GR² is conducted as for ordinary R².

^*Significant at 5%.

^**Significant at 1%.

The estimates in Table A1 indicate that, using production functions with firm-specific fixed effects, we obtain similar estimates of coefficients as in the text for all important purposes. The estimated coefficient of time dummy variables tends to increase. Between COTVEs and PEs and the other types of enterprise group, there is likely to be a difference in the influences of ln(M/K) on the output or productivity, and the estimated coefficient of ln(M/K) × time dummies also tends to increase year by year. Also, there are no large differences between Tables A1 and A2 in the estimated elasticity of output to capital (K) and labour (L).

Table 5

Regression results of estimated firm-specific fixed effects to enterprise-type dummy variables and industry dummy variables^a^,^b

Dependent variable	Estimates of firm-specific fixed effects by equation (5)	Estimates of firm-specific fixed effectsby equation (6)	Estimates of firm-specific fixed effects by equation (7)
	Industry dummy variables
Mining and quarrying	−4.177^**	−4.228^**	−3.499^**
	(−17.583)	(−17.681)	(−15.140)
Foodstuff	−3.285^**	−3.334^**	−2.707^**
	(−34.795)	(−35.080)	(−29.465)
Textile, sewing, leather and furs products	−3.313^**	−3.378^**	−2.714^**
	(−92.811)	(−93.985)	(−78.131)
Other manufacturing	−3.204^**	−3.253^**	−2.604^**
	(−43.131)	(−43.509)	(−36.034)
Production and supply of electric power, steam and hot water	−3.838^**	−4.004^**	−3.393^**
	(−27.133)	(−28.119)	(−24.657)
Coking, gas and petroleum refining	−5.461^**	−5.544^**	−4.811^**
	(−46.183)	(−46.569)	(−41.819)
Chemical industry	−3.212^**	−3.270^**	−2.621^**
	(−71.773)	(−72.568)	(−60.181)
Building materials and non−metal Mineral products	−3.321^**	−3.387^**	−2.746^**
	(−65.606)	(−66.467)	(−55.762)
Metal products	−3.192^**	−3.257^**	−2.581^**
	(−56.694)	(−57.481)	(−47.132)
Machinery and equipment	−3.015^**	−3.080^**	−2.433^**
	(−86.594)	(−87.866)	(−71.805)
Unclassifiable	−3.155^**	−3.214^**	−2.543^**
	(−43.939)	(−44.461)	(−36.397)
	Enterprise-type dummy variables
SOE	−0.574^**	−0.582^**	−0.535^**
	(−16.271)	(−16.393)	(−15.590)
SHE	−0.358^**	−0.349^**	−0.316^**
	(−4.287)	(−4.158)	(−3.895)
PE	0.093	0.013	0.028
	(0.293)	(0.040)	(0.092)
JV	−0.193^**	−0.185^**	−0.155^**
	(−3.527)	(−3.353)	(−2.905)
UCOE	−0.348^**	−0.315^**	−0.312^**
	(−6.884)	(−6.188)	(−6.328)
UCE	−0.381^**	−0.366^**	−0.333^**
	(−6.318)	(−6.033)	(−5.678)
UOE	−0.739	−0.742	−0.628
	(−1.657)	(−1.655)	(−1.449)
Adjusted R²	0.302	0.299	0.301
Number of observations	2018	2018	2018

^aEstimation method is the ordinary least-squares (OLS) method.

^bRegression coefficients. The dependent variable is ln Y (log output). Numbers in parentheses are the t-statistics.

^*Significant at 5%.

^**Significant at 1%.

To measure the differences between the average technical efficiencies of COTVEs and other types of enterprise, the estimates of firm-specific fixed effects obtained in the estimation of production functions (A.1–A.3) were regressed to enterprise-type dummy variables and industry dummy variables. The industry dummy variables are expected to account for the differences in production technologies between industries. The results are shown in Table A2.

In Table A2, the estimated coefficients of the dummy variables for SOEs, SHEs, JVs, UCOEs, UCEs are significantly negative, whereas that of the dummy variable for PEs is never statistically significant. This result indicates that the average technical efficiency of COTVEs has been higher even than that of JVs and SHEs, as well as SOEs, UCOEs and UCEs in the 1990s. This is consistent with the estimates in the text.

It is reasonable to conclude from these results that the estimates of the coefficients in the text are not seriously biased.

Appendix B: Data construction

In constructing the data used for making estimates, the following deflating methods have been used for output and inputs.

deflating method for output

Output is measured by the net output (added value) at 1991 fixed prices, that is,

Open image in new window

where t is the time index, j is the industry index, GV is the gross value of output at current prices, PME represents intermediate inputs at current prices, DEFO is the deflator for output and DEFI is the deflator for intermediate inputs. As indicated above, different deflators are used for the gross value of output and for intermediate inputs; this is called double-deflation. The aim is to avoid biased measurement of real net output arising from differing inflation rates between the final product and intermediate inputs. The price deflator for output is defined as

Open image in new window

where GV90 is the gross value of output at 1990 fixed prices.

Deflating method for intermediate inputs

The intermediate inputs variable (M) is measured by intermediate inputs at 1991 fixed prices (PME_t/DEFI_jt).

The price deflator DEFI for intermediate inputs is derived from the data on ‘the purchase price index of materials, coals, and engines’ and the deflator for ‘the services price indices’ published in the Jiangsu Statistical Yearbook. The DEFI is derived as a weighted average price index. The weights are the average share of each component in the sum of the value of current intermediate inputs for each industry in the sample. The value of the intermediate inputs for each industry are obtained from the input–output table for China for 1992, 1995 and 1997, published in the China Statistical Yearbook. The shares of intermediate inputs for 1991 and 1992 are based on the input–output table for China for 1992; the shares of intermediate inputs for 1993, 1994 and 1995 are based on the input–output table for China for 1995; and the shares of intermediate inputs for 1996 and 1997 are based on the input–output table for China for 1997.

The price deflator was normalised by taking the 1991 index as 100%.

Deflating method for fixed assets

Capital is measured by the real original value of fixed assets. The price deflator for fixed assets is derived from ‘the price indices of investment in fixed assets’ of Jiangsu province in the Jiangsu Statistical Yearbook. This item is divided into three components: ‘construction and installation’, ‘purchase of equipment, tools, and instruments’ and ‘other’. The deflator for fixed assets is derived as a weighted average price index. The weights are the average share of three of each component in the total investment in fixed assets. The weight of each component in the investment differs now not only for each industry but also for each type of enterprise. For example, SOEs have a larger construction share in their fixed assets than other enterprises; this includes more nonproductive resources, such as housing, schools, health facilities and other services for industrial workers and their families. Our procedure therefore includes separate calculations for SOEs, TVEs (including COTVEs) and other types of enterprise, although the Jiangsu Statistical Yearbook did not give data showing the allocation of investment by SOEs, TVEs and other types of enterprise. This Yearbook divides the investment in fixed assets into four types: investment in capital construction, in innovation, in urban collective units and in rural areas. Here, since investments in capital construction and innovation appear to be mainly for use by SOEs, we have assigned the shares of the three relevant components to the sum of investments in capital construction and innovation for SOEs. The share of investment in rural areas is assigned to TVEs, and the share of investment in urban collectives is assigned to the other types of enterprises.

As stated above, capital is derived from the deflated original value of fixed assets, taking the 1991 index as the price base. The figures are as follows:

Open image in new window

where s denotes the rate of physical depreciation of fixed assets in every year, assumed to be 3%, m is the index denoting enterprise type, OF is the nominal value of the original value of fixed assets, DOF is the deflated original value of fixed assets and DEFA is the deflator for the original value of fixed assets.

Finally, labour is taken as the total number of year-end employees.

References

Alchian, A and Demsetz, H . 1972: Production, information cost and economic organization. American Economic Review 62: 777–795.Google Scholar
Basu, S . 1996: Procyclical productivity: Increasing returns or cyclical utilization? Quarterly Journal of Economics 111: 719–751.CrossRefGoogle Scholar
Chang, C and Wang, Y . 1994: The nature of the township–village enterprise. Journal of Comparative Economics 19: 434–452.CrossRefGoogle Scholar
Che, J and Qian, Y . 1998a: Insecure property rights and government ownership of firms. Quarterly Journal of Economics 113: 467–476.CrossRefGoogle Scholar
Che, J and Qian, Y . 1998b: Institutional environment, community government, and corporate governance: Understanding China's township–village enterprises. Journal of Law, Economics, and Organization 14: 1–23.CrossRefGoogle Scholar
Chow, KWC and Fung, MKY . 1998: Ownership structure, lending bias, and liquidity constraints: evidence from Shanghai's manufacturing sector. Journal of Comparative Economics 26: 301–316.CrossRefGoogle Scholar
Dong, X-Y and Putterman, L . 1997: Productivity and organization in China's rural industries: a stochastic frontier analysis. Journal of Comparative Economics 24: 181–201.CrossRefGoogle Scholar
Dong, X-Y and Putterman, L . 2003: Soft budget constraints, social burdens, and labor redundancy in China's state industry. Journal of Comparative Economics 31: 110–133.CrossRefGoogle Scholar
Hsiao, C, Nugent, J, Perrigne, I and Qiu, J . 1998: Share versus residual contracts: The case of Chinese TVEs. Journal of Comparative Economics 26: 317–337.CrossRefGoogle Scholar
Jefferson, GH . 1999: Are China's rural enterprises outperforming state enterprises? Estimating the pure ownership effect. In: Jefferson, GH and Singh, I (eds). Enterprise Reform in China: Ownership, Transition, and Performance. Oxford University Press: New York, pp. 153–170.Google Scholar
Jefferson, GH, Rawski, TG, Wang, L and Zheng, Y . 2000: Ownership, productivity change, and financial performance in Chinese industry. Journal of Comparative Economics 28: 786–813.CrossRefGoogle Scholar
Jiang, C . 2000: Fund resources of township–village enterprises and dynamic change in their financing structure: Analysis and consideration (xiangzhenqiye zijinlaiyuan yu rongzijiegoude dongtaibianhua: fenxi yu sikao). Economic Research Journal (jingji yanjiu) 2: 34–39. (in Chinese).Google Scholar
Jiang, G . 2000: Combination of production operation and capital operation as a way of revitalizing township–village enterprises (shengchan jingying ziben jingying yu xiangzhenqiyede ‘ercichuangye’). Chinese Rural Economy (zhongguo nongcunjingji) 3: 50–55, (in Chinese).Google Scholar
Kornai, J . 1980: Economics of shortage. North-Holland: Amsterdam.Google Scholar
Li, DD . 1996: A theory of ambiguous property rights in transition economies: The case of the Chinese non-state sector. Journal of Comparative Economics 23: 1–19.CrossRefGoogle Scholar
Li, J and Yan, B . 1998: The development patterns of township–village enterprise (xiangzhenqiye fazhan moshi). In: Li, J and Yan, B (eds). Economic Report of China (zhongguo jingji wenti baogao), Economic Daily-Verlag (jingji ribao chubanshe): Beijing, China. pp. 647–665. (in Chinese).Google Scholar
Lu, W . 1999: Agro-economic situation in 1998 and prospects for 1999 (1998 nian nongyexingshi yu 1999 nian zhanwang). In: Ma, G and Wang, M (eds). 1998–1999 Economic Situation and Prospect of China (1998–1999 zhongguo jingjixingshi yu zhanwang). Zhongguofazhan - Verlag (zhongguo fazhan chubanshe): Beijing, China. pp. 19–29. (in Chinese).Google Scholar
Murakami, N, Liu, D and Otsuka, K . 1994: Technical and allocative efficiency among socialist enterprises; the case of the garment industry in China. Journal of Comparative Economics 19: 410–433.CrossRefGoogle Scholar
Pan, C, He, L and Zhuang, Y . 1997: Research on the condition of establishment of collective owned economic structure in our province (wosheng xiangcun jiti jingji zuzhi jianshe zhuangkuang diaocha). Jiangsu Rural Economy (jiangsu nongcun jingji) 7: 13–15. (in Chinese).Google Scholar
Pesaran, MH and Smith, RJ . 1994: A generalized R ² criterion for regression models estimated by the instrumental variables method. Econometrica 62: 705–710.CrossRefGoogle Scholar
Pitt, MM and Putterman, L . 1999: Employment and wage in township, village, and other rural enterprises. In: Jefferson, GH and Singh, I (eds). Enterprise Reform in China: Ownership, Transition, and Performance. Oxford University Press: New York. pp. 197–215.Google Scholar
Shi, X . 1997: Local governments' withdrawing from micro economic agents: A choice across centuries (shequ zhengfu ‘tuichu’ weiguan zhuti -sunan xiangzhenqiye shiji zhi jiao de xuanze-). Jiangnan Review (Jiangnan luntan) 2: 29–30. (in Chinese).Google Scholar
Svejnar, J . 1990: Productive efficiency and employment. In: Byrd, W and Lin, Q (eds). China's Rural Industry: Structure, Development, and Reform. Oxford University Press: New York. pp. 243–254.Google Scholar
Weitzman, M and Xu, C . 1994: Chinese township–village enterprises as vaguely defined cooperatives. Journal of Comparative Economics 18: 121–145.CrossRefGoogle Scholar
Wu, J . 1992: The new features and problems of township and village enterprises in low income areas. Economic Management Research 4: 58–61.Google Scholar
Xu, Z and Zhang, J . 1997: Rapid expansion of fund of township enterprises and decline in the returns–a positive study into township enterprises in Suzhou City, Jiangsu Province (xiangzhenqiye zijinde gaosuzengzhang yu xiaoyi xiahua – jiangsusheng suzhoushi xiangzhenqiye shizhengfenxi). Chinese Rural Economy (zhongguo nongcunjingji) 1: 51–58. (in Chinese).Google Scholar
Yang, Q . 1997: Main contradiction and its solution to continued economic growth in China (woguo jingji shixian chixu zengzhang mianlin de zhuyao maodun yu duice fenxi). China Industrial Economy (zhongguo gongye jingji) 8: 5–9 (in Chinese).Google Scholar
Zou, Y, Dai, L and Sun, J . 1999: A study of township and village enterprises reform in Southern Jiangsu province (sunan xiangzhenqiye gaizhi de sikao). Economic Research Journal (jingji yanjiu) 3: 59–65 (in Chinese).Google Scholar

Copyright information

Authors and Affiliations

Go Yano
- 1
Maho Shiraishi
- 2

1.Faculty of Integrated Arts and Sciences, University of Tokushima, 1-1Minamijosanjima-cho, Tokushima CityJapan
2.Research fellow of the Japan Society for the Promotion of Science, Grad. School of Economics Kyoto UniversityYoshida-honmuchi, Sakyo-ku, KyotoJapan

Efficiency of Chinese Township and Village Enterprises and Property Rights in the 1990s: Case Study of Wuxi

Abstract

Keywords

JEL Classifications

References

Copyright information

Authors and Affiliations

Personalised recommendations

Cite article