1. The four seeds projects covered by this report were approved in the period 1973 to 1978, and closed from 1982 to 1985. They were similar in concept, concerned to ensure that a seed multiplication and marketing system was in place to provide Green Revolution, High Yielding Variety (HYV) seeds to farmers, and were modelled on the Bank's first seeds project, the Tarai Seeds Project (Ln 0614-IN) in Uttar Pradesh. The Bangladesh Cereal Seeds Project (Credit 0410-BAN) was supported by an IDA credit of $7.5 million equivalent approved in June 1973. This was followed by India National Seeds Project (Ln 1273-IN) supported by a Bank loan of $25 million and the Pakistan Seeds Project (Credit 0620-PAK) supported by an IDA credit of $23 million equivalent, both approved in 1976. The final project being evaluated was India Second National Seeds Project (Credit 0816-IN) supported by an IDA credit of $25 million equivalent, and approved in May 1978.
2. The successful Bank-supported Tarai project was taken as a model. With hindsight this can be seen as a good, but second best, design. The option of maximizing the role of a competitive private sector was never seriously considered. Rather the projects took a "systems approach" and included resources for seed breeding, multiplication, processing, marketing and certification. Major investments were in large seed processing plants, plus necessary office, laboratory, communications and transport equipment; accommodation was provided as needed. Technical assistance and significant training components were also included. In India and Pakistan new State (and Provincial) seed corporations were established, and in Bangladesh a new Seed Wing was added to the existing Bangladesh Agricultural Development Corporation (BADC). New seed certification and registration units were also established. As was common at the time, almost all investments were in the public sector.
3. Despite some delays, implementation experience was satisfactory with the exception of Pakistan, where despite Bank approved site engineers provided under technical assistance, two of the four seed processing plants were found to be structurally unsound, and had to be rebuilt (an expense shouldered by the Punjab Provincial Government).
4. At completion project outcome was judged to be at best uncertain, in the case of Pakistan the project was canceled after a two and a half year extension to the surprise and against the wishes of the Borrower. The projects came too late to capitalize on the seed boom associated with the introduction of seed for the first HYVs, and had difficulty in persuading farmers that Certified seed performed better than HYV based seeds saved from the farmer's own fields, and indeed the yield gains, from new seed releases were too small to be readily identified. Thus the projects found themselves demand constrained. In effect the project rationale "there is no point in breeding higher yielding seed if there is no multiplication and marketing system in place" was stood on its head, and the projects found "there is no point in having a seed multiplication and marketing system, if there is no recognizably better seed emerging from seed breeders." 1/ This is too severe a judgment, since farmer saved seed does decline in productivity and become contaminated with weeds over time. Thus even without any increase in yield, a valuable function is performed by feeding good quality Certified seed into the informal seed multiplication system.
Impact: General Perspective
5. Seven years after their PCRs, it is evident that the seeds projects in India and Pakistan used a sophisticated strategy of "creative institutional obsolescence." No less sophisticated for being in large measure unconscious. By promoting state or provincial parastatal seed companies at the expense of a monopolistic national parastatal seed company, the projects mobilized an effective constituency within the public sector in support of decentralization. Not only did this undermine the political and bureaucratic power of the national company, but the decentralized seed companies were less inclined, or less able, to resist the entry of private sector companies. Just as the national seed companies were asked to give way to state companies, so the state companies will be increasingly asked to compete on a level playing field with the private sector, or to give way entirely to the new competition.
6. The original national parastatals made a major contribution by (a) an increase in the physical supply of seed, (b) making farmers, politicians and the agricultural sector in general "seed conscious," and (c) providing an initial cadre of seed scientists and technologists. However, the natural evolution of a monopolistic parastatal is to become increasingly bureaucratic, inefficient, politically powerful and indispensable. The decentralization of the seed industry from a single national monopoly to a series of state or provincial seed companies was thus a brilliant move. It provided a public sector dynamic, including many promotion possibilities, which the central company could not resist. Moreover, it could not be couched in terms of public versus private sector interests. It was an evolution entirely within the public sector.
7. The seed industries have thus evolved from a single parastatal company, to decentralized state or provincial parastatals, to a mixed private-public system, where the public sector supplies predominantly self-pollinated varieties (SPVs), and the private sector supplies hybrids and specialized varieties. Further evolution of the seed industries awaits a re-thinking of seed policy. The project paradigm that there should be a 20 percent replacement rate for SPVs, and the cost should be paid for in full by farmers, may simply be untenable. A reexamination could well lead to a lower replacement rate or a change in the way the subsidy is paid and an enhanced role for the private sector. The argument that the government should be willing to pay the full cost of varietal development and of agricultural extension (because they are public goods), but none of the costs of seed production (even though SPVs are in large measure public goods) does not have intuitive appeal.
8. In like vein, the continued existence of the seed companies brought into existence by the projects, is basically irrelevant to a judgement as to the sustainability of the projects. It is the continued supply of good quality seed to farmers at reasonable prices which is the index of project "sustainability", not which organizations are involved in seed supply. In those cases where parastatal companies evolve to the point where they can compete without differential subsidy with the private sector, their continued existence is to be welcomed. The projects have had a very positive effect on the evolution of the seeds industries in the three countries; that the industries may be in the process of outgrowing some of the institutions which originally fostered their growth is an indication of sustainability, rather than unsustainability.
9. The Cereal Seeds Project in Bangladesh provided the first donor assistance to the seeds sector, a "rock" on which many other donors have subsequently built. The support for the Bangladesh Rice Research Institute was entirely appropriate; and the support of the parastatal Bangladesh Agricultural Development Corporation's (BADC) seed-wing, played a constructive role in the popularization of HYVs. Fortunately the de jure monopoly position conferred on BADC was not reflected de facto due to the self-pollinating characteristics of rice and wheat, which allow farmer multiplication of HYV seeds. Indeed the most widely sown variety "paijam" was imported informally and its multiplication has entirely by-passed the official system. Bangladesh has not benefitted from the competition within the public sector, and hence the expansion of a private sector seed industry has been slower than in India and Pakistan. In the context of a subsequent operation, and in conjunction with other donors, the Bank has prevailed upon the Government to publish a Seeds Policy. Whilst liberalizing the seed industry generally, and allowing both private and public participation in the production and distribution of seeds, this new policy still controls varietal release for five key crops. 2/ Project outcome was judged satisfactory, however less satisfactory than in India and Pakistan where competition within the public sector set the stage for competition with the private sector.
10. In India (as in Pakistan) the projects had clearly had a more beneficial impact at the time of the evaluation than seemed likely at project completion. In addition to the SSCs, there were over 500 private seed companies, 24 of them with links to multinational seed companies, and many of them with their own hybrid development programs. This represents a critical mass of commercial interest and expertise which is almost certainly a new and permanent feature of the agricultural sector in India. A critical mass drawn in large part from seed breeders and other professionals who gained their initial expertise under the auspices of the projects. One of these private seed companies has been successful in pioneering the production and export of hybrid cotton, thus marrying cheap Indian labor with high technology genetics.
11. In Pakistan despite a shaky start, there are now four solid artifacts of the seed project. An active and routine wheat and cotton breeding program, a rigorous seed registration and certification program, a major certified wheat and cotton seed marketing organization, and an extensive cadre of well trained seed industry professionals. The Bank's initial sponsorship of monopoly Provincial seed companies has not prevented the entry of multi-national seed companies and the emergence of some modest private sector national seed companies. The Punjab Seed Corporation having benefitted from five years of able and stable management now operates much as a large farmer-owned cooperative in the United States; whilst breaking-even is important, it does not operate to maximize profits, but rather to maximize benefits to farmers.
12. Key elements of project design, most of which these projects got right, are:
i. Technical Training. The seed industry stretches all the way from genetic research, through varietal development, bulking up, certification, registration, production and marketing. Many of these topics are highly technical, and it follows that a cadre of well trained seed scientists and technologists is a necessary condition for success of the industry. Thus an active training component is an essential element of a good project; technical assistance is a very limited, temporary and imperfect substitute.
ii. Institutional Infrastructure. An initial core of publicly supported genetic research, seed breeding and seed testing is essential for the rapid development of the industry. There is no point in the training component, if trained staff have nowhere to practice their professions. At a later date, there is room for debate and experimentation as to the relative roles of the public and private sectors in research, breeding and certification. However, in developing an industry, the public sector has to stand ready to fulffil any roles not yet satisfactorily performed by the private sector.
iii. Seed Production. Public production and marketing of seed may well be a necessary initial step to (a) demonstrate the existence of a market, and (b) to make seed available pending the development of private sector seed companies. There is every reason such parastatal companies should be encouraged to compete, and if freed from the more onerous of public sector bureaucratic procedures they may be able to do so.
iv. Break-up of Monopoly. It is essential to find some way to avoid the emergence of a single all powerful parastatal monopoly. One alternative may be to foster competing parastatals, another would be to provide for seed importation by companies, other than the seed production parastatal. The apparent confusion due to imbalance in seed supplies from different sources is a small price to pay for avoidance of the policy inertia and bureaucratic procedures which characterize dominance of an industry by a single parastatal, or other forms of central planning.
v. Remove Barriers to Entry. It is important to ensure that there are no unnecessary barriers to entry of private firms into the industry; that government not impose regulations that prevent the private sector from functioning efficiently; and that the private sector have equal access to improved seeds and germ plasm produced from the research system. It is not necessary to have an active pro-private policy stance; but it is essential not to have an anti-private sector stance.
vi. Subsidy. Under no circumstances should a subsidy be used to give differential advantage to the public sector. Any subsidy should be available to all. The administrative problems of doing this, are a significant additional argument against subsidy.
vii. Independent Testing and Certification. National Uniform Varietal Trial open to all; and a highly professional and independent Seed Certification Authority to establish standards for germination, purity, and genetic composition are essential. A truly rigorous Seed Certification Authority is perhaps the most important indicator of a mature seed industry. In India certification is optional, and this has worked well; but farmers need the option of buying certified seed.
viii. Support for Informal System. At least for the next twenty years the bulk of non-hybrid seed planted is likely to be farmer saved. Farmers need increased advice on roguing for off-types, minimizing weed contamination, seed cleaning and storage, and germination testing. 3/
1/ Or, as Ahmad (1994, page 347) says "All efforts to produce quality seed will fail if there is no demand."
2/ Varietal release committees managed by government, also operate in India and Pakistan.
3/ This is being addressed in the Third National Seeds Project for India (Credit 1952-IN).