Rise and decline of the public agricultural systems in China
The severe famine between the late 1950s and the early 1960s in China, and the poor socio-economic situation of the agricultural sector, stimulated a development approach based on modern technologies. The government aimed at the modernization of agriculture by introducing modern varieties (MVs), fertilizer, pesticides, and irrigation. As a consequence, strong national state-run agricultural research, extension and seed systems for MVs were established. The agricultural research and technology development system of CAAS operates at national, provincial, and county levels. Meanwhile, the extension and seed systems are operated by the Ministry of Agriculture (MOA) system, which manages the government’s hierarchy from national to village level. The development and distribution of MVs of cotton and oil seed and the main food crops rice, wheat, and maize have been CAAS’s and MOA’s core task and first priority. These crops are considered decisive for national food security and are therefore placed under the regulation of the state, which controls research and technology development and monopolizes their market distribution. Controlled dissemination of MVs has contributed greatly to China’s technology-driven agricultural growth. The Chinese Green Revolution was considered a success since it served the general goal of the state to achieve national sufficiency in food production. However, the individual needs of farmers, who cultivated the land on a collective basis, were not taken into consideration.
Since the end of the 1970s, the government has gradually introduced reforms to liberalize the state controlled economy. These reforms followed a double-track development strategy by combining the control system on the one hand with free market elements on the other. The government still pursues a technology-driven agricultural development policy with increasing emphasis on biotechnology development and application. However, within the changing socio-economic context, actors of agricultural development are also experiencing reforms. After collective farming was abandoned, farmers began to cultivate crops on an individual household basis and have relatively more freedom to make their own decisions in farming activities. Meanwhile, the dominant public systems, although still state controlled, are undergoing ambiguous changes, as is reflected by the controlled seed system. Officially, every county in every province has a seed company owned and operated by MOA systems. These government seed companies primarily rely on locally produced seed and play a dominant role in seed production and distribution. Only government seed companies are allowed to produce and sell the seed of rice, wheat, maize, cotton and other controlled crops. Active free market trade of seed is limited to vegetables and certain minor crops. After the reform, these local government seed companies have increasingly being privatized and forced to generate income on a commercial basis. However, they still maintain their monopoly at local level.

Economic importance of cotton
Cotton is of superior importance to the Chinese textile industry, which is the largest in the world. This sector employs nine million workers, and its contribution to China’s export volume comprises about 25 per cent of the total. Currently China is the biggest cotton producer in the world; about 50 million farming households grow cotton. Although cotton only occupies between two and three per cent of the total cultivated area, it renders seven to ten per cent of the total value of agriculture.
Since the end of the 1980s, cotton production has decreased due to a decline in both yield and coverage area. The decline in yield of 15 to 30 per cent has mainly been caused by bollworm infestation. In 1992 and 1993, outbreaks of bollworm infestation in China caused direct economic losses of about US$ 630 million. Furthermore, farmers were discouraged from growing cotton by the controlled and relatively lower price of cotton compared with non-controlled cash crops like vegetables and fruit trees. As a result, the national growing area decreased by 10 to 15 per cent, and there is a tendency for cotton production to move from relatively favourable areas towards more marginal regions. For instance, the cotton growing area of Hebei province, a more favourable area, decreased from over 0.67 million hectares before 1990 by 70 per cent to 0.2 million hectares after 1990. However, the cotton growing area in Xinjian, a marginal area, increased by about 20 per cent during the same period.

Development of Caas Bt cotton
Cotton is the crop to which most insecticides are applied in the world. Therefore, insect resistance in cotton has become a target of modern biotechnology research. This led to the development of insect resistant crops through the insertion of a Bacillus thuringiensis (Bt) gene. This gene causes the production in plant tissues of crystal proteins that are toxic to certain insects. This gives plants protection against specific insect pests without pesticide application (see also Monitor No. 29). Cotton was among the first crops for which the biotechnological introduction of the Bt gene was commercialized.
In the face of the serious bollworm damage, scientists at CAAS began their research on Bt cotton in 1991. In 1993, Bt cotton research and technology development were officially included in the 863 High Technology R&D Programme, which is the national programme that receives priority status in the government’s planning. In 1994, China became the second country in the world after the USA in which Bt cotton had successfully been field-tested. CAAS’s Bt cotton was tested in eight provinces including both favourable and unfavourable cotton growing areas. By the end of 1996 CAAS had developed two Bt cotton varieties and one of them, with its variety name Guokang, gained the state’s permission for commercial introduction in four provinces. However, until 1997 CAAS’s Bt cotton covered only about 667 hectares. Although the coverage increased to 13,333 hectares in 1998, the newly developed transgenic cotton has been adopted much more slowly than intended.

Difficulties in commercializing CAAS Bt cotton
According to several Chinese researchers and officials, CAAS’s Bt cotton, Guokang, can compete with Monsanto’s Bollgard Bt cotton in terms of insect resistance. Monsanto began work on Bt cotton in the USA in 1987 and commercially introduced Bollgard in 1996. In 1998, cotton with Monsanto’s Bt technology covered about 40 per cent of the total cotton area in the USA, and at present Monsanto is the most important player in Bt cotton development and distribution worldwide.
However, agronomically both cotton varieties have their own strengths and weaknesses. Bollgard cotton has a more stable agronomic performance in favourable production environments with plastic mulching and irrigation. Meanwhile, CAAS Bt cotton is more adaptable to small-scale farming systems with relatively poor conditions, and is higher yielding than Bollgard in unfavourable areas.
The distribution of CAAS’s Bt cotton seems to have suffered from institutional and policy constraints such as:

• Decreasing financial resources. The government’s reduction of funding and privatization of the public system means that public institutions and their staff increasingly have to generate income through commercial activities such as hybrid seed multiplication, or trading of seed, fertilizers, and pesticides. However, these activities are detrimental to their research mandate.
• Lack of financial incentives for plant breeders and scientists. At present, plant breeders and research institutes that develop most of the new varieties get paid only a nominal fee from seed companies for providing them with parent seed. However, the seed companies reap most of the benefits by the commercialization of these newly developed varieties. This system creates a strong disincentive for breeders and researchers. Furthermore, a highly qualified plant breeder earns approximately the same salary as a manual worker. As a consequence, most of the promising young scientists in CAAS, who initiated the Bt cotton research, have left the institution, mainly due to the poor working conditions and the irrational compensation system.
• Outdated institutional framework with poor linkages. CAAS has suffered from linkage and cooperation deficits between biotechnology research, variety selection and improvement, and seed production and distribution. Most of the characteristics that were identified as useful by molecular biologists were not rapidly incorporated into appropriate commercial crop varieties. Furthermore, it was even more difficult for these traits to be distributed through the public seed production and extension system at local levels and amongst farmers. The top-down approach of linear technology design, development and diffusion that was successful in the Chinese Green Revolution fails to work efficiently under current circumstances in which demands of farmers and markets play a more prominent role.

Entry of Monsanto’s Bt cotton
Monsanto had been trying to enter the Chinese market since 1993. At the end of 1997 the company received the national government’s permission to commercialize its Bollgard Cotton in Hebei. Local newspapers reported that, already in its first season of planting in 1998, Monsanto’s Bollgard cotton was cultivated on approximately 80,000 hectares. According to the official provincial statistics, this variety was covering about 40 per cent of all cotton growing area in the province. How did Monsanto manage to enter the controlled cotton seed market in Hebei?
According to Wang Shikui, deputy director of the Department of Science and Technology in Hebei, in 1994 Monsanto started negotiations with the provincial government for a collaborative distribution of Bt cotton in Hebei. In 1995, the government imposed a comparative field test of three Monsanto Bt cotton varieties and two local non-Bt varieties in three testing areas. The major testing criteria were insect resistance without the use of insecticides, and yield. Not surprisingly, one of the Monsanto Bt cotton varieties, with the name 33B, yielded 30 per cent more than the local ones. Based on this result, the provincial government decided to collaborate with Monsanto in producing and distributing Bollgard cotton through the government seed system and extension mechanism. With the result of the field trial, and support from the provincial authorities, Monsanto finally gained approval also from the national government in 1997.
Foreign investors can produce cottonseeds in China as long as they form a joint venture with a Chinese company. Therefore, in November 1996, a joint enterprise named JiDai, was established between the Hebei provincial government seed company, Monsanto, and the US cotton seed producer Delta & Pine Land Co. (D&PL). In 1998, D&PL was acquired by Monsanto. The total investment of JiDai is US$ 8.4 million of which Monsanto and D&PL jointly own 67 per cent shares and the Hebei provincial partner owns 33 per cent. According to Pei Jianzhong, General Manager of JiDai and former deputy director of the Hebei provincial seed company, JiDai practices an "introduction from outside and production inside" system. This means that Monsanto introduces the foundation seed including its Bt technology, but seed production, processing and distribution are all operated locally by JiDai. Obviously, JiDai as a partly government-owned seed company has access to the entire government seed system, extension service, and marketing system. It uses county government seed companies as its sales stations and employs the local government officials and extension workers as salespersons. There are more than 5000 retailers in most of the cotton growing counties in Hebei, comprising a complete marketing network. Contracts between Monsanto and JiDai, and between JiDai and the salespersons, determine that the latter are obliged to distribute Monsanto’s Bt cottonseed exclusively. Furthermore, since these retailers are local officials, they are allowed to use government intervention measures in the distribution of seeds to guarantee that farmers fulfil their quota.
Monsanto’s entry into the Chinese market has created great debate and controversy among Chinese agricultural policy makers and scientists. Some of them argue that the central government should protect the market by re-establishing monopoly seed production and distribution, whereas others consider competition as helpful to the transformation of CAAS in particular and economic development in general.

Emerging issues for the public sector
The previously mentioned constraints of the public institutions alone are not fully sufficient to explain the difficulties of CAAS in commercializing its Bt cotton. More important is the combination of these limitations together with CAAS’s loss of control over the local government system. It can be concluded that in China the institutional reforms of the agricultural systems do not parallel the sweeping economic reform, and rapid development of the market economy. The centralized public agricultural system, which was set up in the context of the Green Revolution, is inappropriate and inefficient in the new social and economic setting of market liberalization. Inconsistent reforms resulted in institutional conflicts, and gaps leading to the constraints and obstacles for CAAS to develop and distribute its technology efficiently.
In Hebei province, this gap in the transformation process worked to the advantage of Monsanto. The company successfully entered the market by gaining access to local government systems and by using the government’s monopoly in seed production and distribution. To achieve this goal, Monsanto could rely on its superior financial resources, its marketing knowledge, and efficient management, which in the end gave it a competitive edge over CAAS.
Some preliminary conclusions from this comparative analysis can be of relevance for the future of agricultural research and biotechnology deployment for policy makers and scientists in China:

• Public institutions should adjust according to the free market reforms. Maintenance of monopoly under the new system is untenable and results in poor management of public agricultural research, product development and marketing. Having a private monopoly instead of a public one is not a satisfying alternative neither for national development nor for opening new options to farmers.
• A public institution undergoing transition towards privatization should not be left alone at the mercy of free market forces. The government should have a strategic plan and corresponding logistical support to assist these institutions in a smoother transition.
• The Chinese government should draw out clear policies on the role of public institutions in technology development and distribution vis-à-vis the government’s national goals and vis-à-vis the relationship with the private sector. However, farmers in marginalized areas might still depend on public technology. If interaction with these farmers could lead CAAS to take their needs and interests into account, this could become CAAS’s ‘comparative advantage’ in technology development while still serving the general mandate of the public system. Evidently, this would be a decision on political grounds, which does not solve the problem of income generation for public institutions.

Department of Communication and Innovation Studies, Wageningen Agricultural University, Hollandseweg 1, 6706 KN Wageningen, the Netherlands.
Phone (+31) 317 483 297; Fax (+31) 317 484 791; E-mail yiching.song@alg.vlk.wau.nl

This article is the result of the scholarship programme of the Biotechnology and Development Monitor

Sources
Krattiger, A.F. (1997), Insect Resistance in Crops: A Case Study of Bacillus thuringiensis (Bt) and its Transfer to Developing Countries. ISAAA Briefs No. 2. Ithaca, NY, USA: ISAAA.

James, C. (1998), Global Review of Commercialized Transgenic Crops: 1998. ISAAA Briefs No. 8. Ithaca, NY, USA: ISAAA.

Lin, J.Y. (1998), How did China Feed itself in the Past? How will China Feed Itself in the Future? CIMMYT Economics Programme, Second Distinguished Economist Lecture. Mexico City, Mexico: CIMMYT.

Eisa, H.M. et al. (1994), Cotton Production prospects for the Decade to 2005: A Global Overview. World Bank Technical Paper No. 231. Washington, D.C., USA: World Bank.

Song, Y. (1998), "New" Seed In "old" China; Impact Study of CIMMYT’s Collaborative Programme on Maize Breeding in Southwest China. PhD Thesis, Wageningen Agricultural University, the Netherlands.

Personal communications with Q. Wang, S.R. Jia (Biotechnology Research Centre of CAAS), W. Shikui (Hebei provincial government),
L. Junlan, Z. Xiangyun, Z. Degai, L. Yongzeng (Hebei Cotton Research Institute), P.J. Zhong (JiDai ), and C. Martin (Monsanto China).