## Competitive markets for seeds of GM local varieties

Assuming that the fixed cost of introducing the technologies are covered by the public sector, competitive seed sellers will charge farmers V- per land unit of GM seeds. The GMVs may be either fully or partially adopted. In the case of partial adoption, some land will continue to be grown with traditional variety j. Let A'- be total acreage of variety j (traditional and

GM), and let Ajl be the GM acreage. The acreage of the traditional variety j will be equal to A' - Ajl. There are three possible outcomes under competition:

• Cjh No adoption when PMp'J (ff)- PMp](()) <V,. In this case, Af = 0, A) = A?.

• CJ2: Partial adoption when PMpj(0)-PMplj(o)>Vj > PM/?7(A;0)-PMp7(A°). Here A^A® where PMpf(Af) = PMP%A?) +V, and A) = A*.

• CJ3: Full adoption when PMPJ[a°j)-PMp^A^Vj. A) = Af when PMpJ (a™1 ) = MCj + Vj.

The assumption that the Mp gap between the GMV and traditional variety declines with acreage is the key to the above results. If gains from adoption cannot cover the variable cost even for the first acre, adoption will not occur. If marginal gains from adoption are greater than the variable cost, but only for a subset of the acreage, there will be partial adoption; and if the marginal gains from adoption are greater than the variable cost for all acres, there will be full adoption and the acreage under the crop might even increase (recall that we have assumed the total output market is sufficiently large that this change has no impact on output prices).

Fig. 14-la depicts case C™ 1 with no adoption when PMp°j(o) + Vj >

PMpf(0). Fig. 14-lb depicts case Cj 2 of partial adoption, and Fig. 14-lc depicts case CJ3 with full adoption. These figures demonstrate that high variable costs may discourage adoption entirely, while reduction in these costs may result in partial or, eventually, full adoption. In the figures, the same notation should be used as in the text, that is, replace Mb by PMp.

The introduction of GMVs affects output and net benefits of land utilized with variety /. The output when GMVj is available becomes

if cf 1

if Cf 2

if Cf 3

The net change in social benefit is

ml j

and it represents the difference between the production gain and the variable and fixed cost of the new variety.

Since up until now almost all crop biotechnologies have been developed and commercialized by private companies in monopoly situations, introduction of GMVs by the public sector has rarely occurred in reality. However, the case of Bt cotton in China closely corresponds to this scenario. Although Monsanto and Delta and Pine Land (D&PL) have introduced U.S. Bt cotton varieties in China, the Chinese Academy of Agricultural Sciences has developed and commercialized its own Bt cotton technology, which can be freely used by public and private sector breeders. Due to weak IPR protection, the Monsanto technology also has been incorporated into Chinese cotton varieties by local organizations, without payment to the company.

At present, there are 22 officially registered local Bt varieties and five imported ones available on the market (Pray et al., 2002). Adoption of Bt varieties has occurred in approximately 35% of the Chinese cotton sector and is increasing rapidly. There is no indication that GM technology has a negative effect on cotton biodiversity. On the contrary, the Bt varieties imported from the United States appear to have broadened the local germplasm base.

A similar situation could occur in other countries that do not protect IPRs but have a strong breeding capacity. If foreign GMVs are introduced in these countries, breeders can freely use these varieties to cross-breed the transgenic traits into their own germplasm. The trade-off, however, is that without IPR enforcement, private seed industry development is hampered and technology transfer from abroad is discouraged.

0 0