Agriculture has been an important engine of economic development, and the mainspring of economic progress in agriculture has been productivity improvements driven by technological change that is fueled by research and development (R&D). Since World War II, agricultural productivity has more than doubled in the United States, as in many other countries. California agriculture today produces more than twice the output of 1950, using roughly the same total input — although with less labor and land, and more capital and purchased inputs.

These gains can be attributed to new biological, mechanical and chemical technologies, including improved genetic material, machines, fertilizers and pesticides, and knowledge. The current wave of technological progress continues this pattern, while emphasizing information technologies and biotechnology — in particular genetically modified (GM) crops. For many, GM crops represent the hope for a future with less hunger and malnutrition, and for a more sustainable agriculture with more varied, cheaper and safer food. For others they are cause for serious concern about the environment and food safety. Regardless of how we may feel about it, the juggernaut of technological change continues and the biotechnology revolution is well under way in the United States and other countries.

The challenge for public policy is to determine what regulations should be applied to govern the development and use of these technologies, and what other types of intervention may be necessary, such as public investments in research to correct for private-sector underinvestment. In the case of horticulture — the cultivation of fruits and vegetables, tree fruits and nuts, turfgrass, flowers and ornamental crops — these issues are sharply drawn because the private sector has not found it profitable to develop or commercialize many GM crops in the current political, legal and market environment.

What will happen in biotechnology applied to horticultural crops is up to the government, for a variety of economic reasons. Some of these aspects may be unique to GM horticultural crops but many are common to GM crops generally, and similar issues arise with some new non-GM technologies.

Public, private roles in ag science

Without government intervention, the rate of innovation will be too slow, reflecting both underinvestment in research and under-adoption of some research results. Both problems are related to the nature of property rights for research results. “Free-rider problems” occur when property rights are incomplete, and private investors can capture only part of the returns to their investments in certain types of research (such as developing new crop varieties); as a result, their incentives to invest are reduced. On the other hand, when the rights to research results are protected, such as by patents or trade secrets, the owner of a new variety can charge monopoly prices, unduly limiting the use of that variety. Intellectual property rights (IPRs) are a double-edged sword: to the extent that they provide a greater incentive for investing in research they are also likely to result in lower adoption rates.

Governments have addressed the incentive problems in agricultural research in several ways. Federal and state governments (as well as industry) have funded agricultural research at public institutions such as the U.S. Department of Agriculture (USDA) and state agricultural experiment stations associated with land-grant colleges. This approach allows an increase in total research without the problems associated with monopoly pricing of inventions. However, even though the investment has paid handsome dividends, it is increasingly difficult to sustain the past levels of funding for public agricultural R&D, in the face of general budget problems and declining political support for public science funding, including agricultural science (Alston et al. 2000). Governments have also acted to strengthen IPRs applied to plants and animals as well as mechanical technologies; and changes in IPRs, especially in the 1980s, were crucial for the agricultural biotechnology development that followed. Partly as a reflection of enhanced IPRs, in the United States, private-sector funding of agricultural research has been growing faster than public-sector funding and now exceeds it.

The balance in agricultural R&D between the private and public sectors varies among types of research. For instance, until recently the private sector emphasized agricultural R&D pertaining to mechanical and chemical technologies, especially pesticides, where IPRs are effective; and the government was more important in other areas such as improving crop varieties. Private involvement was dominant in cropvariety research only in hybrid corn, where the returns were well protected by technical restrictions on copying or reusing saved seed, trade secrets and other legal rights. Changes in the institutional environment and the form of new IPRs, combined with new scientific possibilities associated with modern biotechnology, resulted in a shift in the private-public balance in research to improve crop varieties. As the balance shifts toward private research, new attention must be paid to old questions about whether the private investment in crop research will be sufficient, whether the allocation of those resources (say, among crops) will be optimal, whether the results will be adopted rapidly and widely, and what role the government should play.

Economic and market aspects

The development of new technologies through R&D is only one element of the picture. The technologies must also be approved for commercial application and economically attractive enough to be adopted by farmers. The experience with other biotech crops has lessons for horticultural biotechnology.

Biotech crops have been a commercial reality only for a few years but they have made very rapid inroads in some parts of the market. In particular, pest-resistant and herbicide-tolerant corn, soybeans, canola and cotton were rapidly developed and adopted in the United States and to a lesser extent in other countries (James 2000). To date, the successful GM crop varieties have emphasized “input traits,” related to reducing the use of chemical pesticides or making them more effective, rather than “output traits,” related to product quality. Why has there been rapid development and adoption of GM cropping technologies for these crops and not other important crops, such as wheat and rice? The likely reasons relate to the nature of supply and demand for new technology, and the economics of adoption.

Benefits to farmers and others.

The total benefits from farmers adopting any new cropping technology are approximately equal to the benefits per acre times the number of acres affected. With pest-resistant crop varieties, these benefits come primarily from reduced costs for applying chemical pesticides and increased yields, after an allowance for regulatory requirements for refugia to manage resistance. The distribution of these total benefits between farmers (and ultimately food and fiber consumers) on the one hand, and the technology suppliers on the other, is determined by the size of the premium charged for the use of the new technology, but this premium also affects the incentives for farmers to adopt the technology.

Economic studies suggest that farmers and biotech companies have shared in the benefits of biotech crops and that the net benefits have been large. Gianessi et al. (2002) conducted 40 detailed U.S. case studies of biotech cultivars. They estimated that in 2001, eight biotech cultivars adopted by U.S. growers provided a net value of $1.5 billion to growers, reflecting increased crop values and cost savings. They further estimated that the 32 other case-study cultivars would have generated an additional $1 billion in benefits to growers if they had been adopted, bringing the total potential benefit in 2001 to $2.5 billion. Of this annual total, the lion's share was for herbicide-tolerant crops ($1.5 billion per year), followed by insect-resistant crops ($370 million per year). These estimates do not represent the total economic impact because the geographic analysis was limited in scope, and they do not include any benefits to the seed companies and biotech firms that produced the technology.

Environmental concerns.

Private benefits and costs from biotech crops accrue to growers and consumers of the products, along with seed companies and biotech firms. If the new technology involves environmental risks (as some fear may be the case with biotech crops) or replaces technology that involves environmental risks, there will be additional environmental costs and benefits to take into account as an element of national costs and benefits. For instance, pest-resistant crops can reduce the application of broad-spectrum chemical pesticides, which are hazardous to farmworkers, compromise food safety and impose a burden on the environment. The economic studies to date have not assessed these environmental costs and benefits. However, Gianessi et al. (2002) estimated that adoption of the eight current cultivars allowed a reduction in pesticide use of 46 million pounds in 2001, and the 32 potential cultivars could have allowed a further reduction of 117 million pounds. The relevant comparison then is between the environmental risks associated with these biotech crops and those associated with the annual burden on the environment of 163 million pounds of chemical pesticides that could be avoided by growing biotech crops instead - 66 million pounds in California alone, where 185.5 million pounds of pesticides were used in 1999 (Mullen et al. 2003)

Market acceptance.

On the demand side, farmers will adopt biotech varieties if the perceived net benefits to them are large enough, and this depends on the perceived market acceptance of GM crops. Concerns have been raised about the possibility that GM crops may be unsafe for consumers because of allergens or other, as yet unidentified risk factors, about risks to the environment and to the economy from uncontrolled genetic drift, and about the moral ethics of tampering with nature. The GM varieties that have been developed and adopted extensively to date have not experienced significant price discounts because of buyer resistance. This can probably be attributed to the nature of the crops. For feed grains, the buyers are other farmers who are comfortable with the technology, and for fiber crops such as cotton the food safety concerns do not apply. For the major food grains, wheat and rice, even if the farm-production economics potential of GM varieties is as large as for feed grains, market acceptance may differ sufficiently to limit their adoption. Rather than another farmer, the relevant buyer for these crops is a food processor, manufacturer or retailer who may be reluctant to risk negative publicity or to risk losing consumers who would prefer a biotech-free label or who may not be confident that the biotech and nonbiotech grain can be segregated.

Processors and retailers.

It is not sufficient that farmers and consumers perceive net benefits from GM crop varieties. The adoption of biotechnology must provide net benefits to other participants in the marketing chain, such as food processors and retailers. Pricing of the technology may be a critical factor. Even if the new technology is more cost-effective than the traditional alternative, monopolistic pricing could mean that the technology supplier retains a large share of the benefits. The cost savings passed on to processors and consumers may be a small fraction of the total benefits, rendering incentives for processors, retailers and consumers to accept the technology comparatively small. Processors and retailers can effectively block a new technology if it does not clearly benefit them, even if there would be net benefits to the general public.

Fixed costs.

The size of the market matters. The cost to develop a new variety is essentially the same whether it is adopted on one acre or a million acres, but the benefits are directly proportional to the number of acres on which the variety is adopted. This is why biotech companies have focused on developing technologies for more widely planted agronomic crops, especially feed-grain and fiber crops for which market barriers are lower.

The technology developer must also obtain regulatory approvals. It is difficult to obtain precise information on costs of regulatory approval for biotech crops and chemical pesticides, but according to available estimates, the total cost of R&D — from “discovery” to commercial release of a single new pesticide or herbicide product — exceeds $100 million, and regulatory approval alone costs more than $10 million. A new technology must generate enough revenue for the developer over its lifetime to cover these costs, and for some crops the total acreage is simply not sufficient. Given the large fixed costs associated with regulatory approvals for specific uses, agricultural chemical companies have concluded that the potential market is too small to warrant the development of pesticides for many of California's specialty crops, which have become technological orphans.

It does not follow that the government should invest in developing new conventional or GM pest-control technologies for these orphan crops. If the current regulatory policy and process is appropriate and efficiently implemented then the high cost is not excessive; if a new technology cannot generate benefits sufficient to pay those costs, then it is simply not economical to develop that technology. The question for technology orphan crops is whether it is possible to reduce the costs of R&D and regulatory approval sufficiently to make it profitable for the nation and the private sector to change their orphan status.

Markets for horticultural biotech

On the supply side, “horticulture” includes an enormous diversity of fruit and vegetable crops, but it also includes many nonfood species, such as ornamentals, flowers and recreational turfgrass. Collectively these horticultural crops compare well with major agronomic crops in terms of total value in the United States. However, they use much less acreage, and the market size for some biotech products depends on both acreage and production value. In 2000, the United States produced fruits, nuts and vegetables with a total value of more than $28 billion, of which California contributed about $14 billion (table 1). In addition, horticulture includes a small number of larger-scale crops (such as potatoes and onions, apples and wine grapes) as well as a large number of smaller-scale crops (such as Brussels sprouts and persimmons). At current costs for R&D and regulatory approval, it is unlikely that biotechnology products will be developed and achieve market acceptance for many of these smaller-scale crops in the near future ( see sidebar, page 84 ). Further, experimentation with perennials such as grapes, nuts and fruit trees is comparatively expensive (because the experimental unit is larger and takes more time), and it is costly to bring new acreage into production or replace an existing vineyard or orchard with a new variety.

On the demand side, the market for horticultural products, especially fresh fruits and vegetables, is undergoing important changes associated with the changing structure of the global food industry ( see sidebar, page 82 ). Increasingly fewer and larger supermarket chains have been taking over the global market for fruits and vegetables, especially fresh produce, and changing the way these products are marketed. Because fresh produce is perishable and subject to fluctuations in availability, quality and price, it presents special problems for supermarket managers compared with packaged goods. Supply-chain management, and the increasing use of contracts that specify production parameters as well as characteristics and price, is replacing spot markets for many fresh products. A desire for standardized products, regardless of where they are sourced around the world, could limit the development and adoption of products targeting smaller market segments, unless retailers perceive benefits and provide shelf space for diversified products — such as biotech and nonbiotech varieties of particular fruits and vegetables.

On the other hand, an increasingly wealthy and discriminating consuming public can be expected to continue to demand increasingly differentiated products — with an ever-evolving list of characteristics such as organic, low-fat, low-carbohydrate and farm-fresh. Hence retailers will have to balance the cost savings and convenience associated with global standardization against the benefits from providing a greater range of products, which will include GM products when retailers begin to perceive benefits from stocking them. Unlike other types of foods, fruits and vegetables are often consumed fresh and in clearly identifiable and recognizable form. This has implications for perceptions of quality and food safety that may influence consumer acceptance — perhaps favorably, for instance, if a genetically modified sweet-corn could be marketed as reduced-pesticide ( see page 99 ).

Other elements of GM horticulture — such as nonfood products, ornamentals or turfgrass — have advantages in terms of potential market acceptance. GM trap crops, which provide pesticide protection for other crops, and GM sentinel crops, which signal the presence of pests or provide other agronomic indicators — may be used in food production without overcoming barriers of acceptability to market middlemen or consumers ( see page 89 ). Biotechnology products designed for home gardeners may be more readily accepted because the grower is the final consumer.

Market acceptance in the United States is also linked to continued access to export markets, particularly in the European Union and Japan where restrictions have been applied to biotech foods. The relative importance of the domestic market could help to account for the success of the GM feed-grain technologies in the United States, and it may also help to account for the success of these and other GM technologies in China. China is comparatively important in horticultural biotechnology — its investment in agricultural biotechnology is second only to the United States, but with a different emphasis, including significant investment in horticultural biotechnology ( see page 112 ).

Implications for government policy

The technological potential for GM horticultural crops appears great, particularly when we look beyond the “input” traits that have dominated commercial applications to date, to opportunities in “output” traits, such as pharmaceuticals and shelf-life enhancements. Because delays in socially beneficial technologies mean forgone benefits, there may be a legitimate role for the government in facilitating a faster rate of development and adoption of horticultural biotechnology products. For instance, the government could reform property-rights institutions to increase efficiency and reduce R&D costs. IPRs apply to research processes as well as products, and limited access to enabling technology or simply the high cost of identifying all of the relevant parties and negotiating with them, may be retarding some lines of research — a type of technological gridlock (Binenbaum et al. 2003). Nottenburg et al. (2002) suggest a government role in improving access to enabling technologies. Similarly, the government could revise its regulations to increase efficiency and reduce costs for regulatory approvals. Instead of requiring a completely separate approval for each genetic transformation “event,” it may be feasible to approve classes of technologies with more modest specific requirements for individual varieties.

The government could also reduce some barriers to adoption, especially market acceptance of biotech food products, by providing information about their food safety and environmental implications. The biotech industry and agriculture can have an influence here, too. The general education of consumers and market intermediaries about biotech products may be facilitated in a process of learning by experience with products — such as nonfood applications, or home garden applications — that have good odds of near-term success because of low barriers to market acceptance and good total benefits.