Cover Image
close this book Assessing the Impacts of Agricultural Biotechnologies
View the document Foreword
View the document Preface
View the document Acknowledgments
View the document Executive Summary
Open this folder and view contents Review of Previous Studies
Open this folder and view contents Methodological Tools and Approaches
Open this folder and view contents Integrating Impact Assessment Data into Decision-Making
View the document Participants

Executive Summary

Brent Herbert-Copley

Senior Program Officer, IDRC, Ottawa, Canada

Introduction

The workshop on which this volume is based represents one of the first formal activities of the "Canada–Latin America Initiative on Biotechnology, the Environment and Sustainable Development" (CamBioTec). The Initiative, launched by IDRC in January 1995, is an ambitious program designed to promote the application of biotechnology-based products and applications to respond to critical needs in the agrifood and environmental management sectors of selected Latin American studies. This is to be achieved by supporting an integrated set of activities designed to identify opportunities for biotechnological innovations, and to overcome key bottlenecks to their effective application.

The decision by IDRC to host this workshop reflects a recognition of the need for careful, rigorous analysis of the social, economic, and environmental impacts of agricultural biotechnology applications. From the outset, it was felt that this should be a crucial part of the overall work program of the initiative, an essential complement to activities geared to promoting the transfer and application of specific biotechnology-based products, and improving the capacities of biotechnology-based enterprises.

Understanding the way in which biotechnology applications affect socioeconomic and environmental variables is not simply an academic exercise. Impact assessment data is important to a range of decision-makers - from public sector research agencies involved in supporting biotechnology research; to regulatory bodies charged with granting approvals for the introduction of specific products; to a range of private-sector and nongovernmental actors faced with decisions regarding future investments in the adoption and application of biotechnology-based products.

Unfortunately, the very diversity of interests at stake complicates the task of assessing the impacts of particular biotechnology applications. As one of the participants in the workshop noted, the various constituencies for impact analysis approach the issue from differing perspectives and with differing needs in terms of the type of information and level of detail they require. The point of view and methodological approach of a university-based sociologist are likely to differ widely from those of a regulatory official, or from a representative of a biotechnology company attempting to gauge the market for a new product.

Even where there is agreement over the objectives and uses of impact evaluation, there is still a great degree of uncertainty about just how to organize and carry out such research. A cursory look at any of a number of recent reviews of the literature in this field confirms this fact. Thus, for example, a 1991 survey by Martin Fransman of the University of Edinburgh concluded as follows:

One of the notable facts to emerge from the present survey is the extreme scarcity of rigorous studies analyzing the economic and social effects of biotechnology in advanced industrialized and Third World countries. While studies on the use of biotechnology in various application areas and countries abound, there are very few good studies that examine effects. In part this is due to the complexities that are an inherent part of any rigorous study of effects.... Again, the task that lies ahead is in significant measure one of refining approaches and methodologies (Fransman 1991, p. 75).

More recently, Joel Cohen of the Intermediate Biotechnology Service (IBS) has noted that:

Systematic evaluation of the socioeconomic impacts [of agricultural biotechnology] in developing countries is therefore still in its infancy. Also considerable variation exists among economic models for impact evaluation so that any prediction of socioeconomic impact of biotechnology in developing countries must be regarded with great caution (Cohen 1994, p. 31).

To be fair, the volume and quality of work in this field has improved considerably in the past 4–5 years. But, at the same time, the need for impact assessment has also grown, as the pace of introduction of biotechnology-based products has begun to quicken. In addition, there is increasing concern to include an analysis of the environmental impacts of agricultural biotechnology applications alongside socioeconomic analysis, adding yet another wrinkle to the already complicated task of impact assessment.

Background and Objectives of the Workshop

With these factors in mind, IDRC strongly felt that any attempt to launch further research in the field of biotechnology impact assessment should begin with a careful discussion of methodological approaches and tools. The workshop, held at the IDRC offices in Ottawa from 16 to 17 May 1995, was designed as a preliminary discussion among a select group of experts from Canada, Latin American countries, and international agencies involved in the field of agricultural biotechnology. Its objectives were:

· To identify different methodological approaches for the assessment of socioeconomic and environmental impacts of agricultural biotechnologies, as well as to review the potential and limitations of methodologies applied in previous studies;

· To recommend approaches and methodologies for impact assessment that could be tested in pilot studies to be sponsored by IDRC as part of the CamBioTec initiative; and,

· To identify specific opportunities for such pilot studies and, where possible, to suggest researchers and research institutions working in this field who could carry out such studies.

Emphasis was placed on fostering exchange among the participants rather than on a series of lengthy formal presentations. The first day of the seminar was organized in three sessions, each kicked off by two-three short presentations: the first session concentrated on presenting the results of existing impact evaluation studies, to provide a common frame of reference for subsequent discussions; the second session focused on discussion of alternative methodological tools and approaches for impact assessment; and, the final session of the day looked at the ways in which impact assessment results could be integrated into decision-making processes. The second day of the workshop consisted of working group discussions of possible approaches to be applied in pilot studies funded under the CamBioTec program.

The next section of this chapter provides a brief summary of each of these sessions, outlining the key points made in formal presentations and in the ensuing discussions. The final section then outlines some of the general conclusions emerging from the workshop.

Discussion at the workshop was lively and wide-ranging, which reflects the diverse nature of the participants. In organizing the workshop, an effort was made to include participants from a variety of geographic locations and professional backgrounds. The participants included social scientists from a number of disciplines (economics, agricultural economics, sociology), agricultural specialists, biotechnology researchers, and individuals with a background in the study of innovation policy and management.

The majority of the participants were currently or had been previously involved in carrying out impact assessment studies, although some also represented institutions supporting or carrying out biotechnology research, or public sector agencies involved in the regulation of biotechnology products. Although this diversity of viewpoints at times made it difficult to arrive at a consensus, it was considered essential to deal with the range of issues and interests at stake in the evaluation of the impacts of agricultural biotechnology. As will be seen in later sections of this chapter, the discussions at the workshop reiterated the importance of this kind of "multistakeholder" approach and suggested that efforts to assess the impacts of agricultural biotechnologies should involve the broadest possible consultation with affected parties.

Summary of Discussions

Review of Previous Studies

The first set of papers presented at the workshop were designed to set the stage for the subsequent discussions by providing an overview of some of the conclusions emerging from existing studies of the impact of agricultural biotechnologies in the Third World and by reviewing the methodological approaches followed in such studies.

Any such review, of course, is of necessity partial and cannot do justice to the various studies that have been carried out. The three papers presented here, however, provide a good indication of the range of existing work on the subject. They also illustrate the diversity of possible levels of analysis - from the broad impacts of biotechnological advances on patterns of international production and trade in the global economy, to in-depth analysis of the impacts of specific biotechnology applications on development at the national and local levels.

The paper by Carliene Brenner reviews the results of a series of studies on biotechnology and agriculture in the Third World, carried out since the late 1980s as part of the Organisation for Economic Co-operation and Development (OECD) Development Centre's research program on technological change in developing-country agriculture. This has included a study of the possible impacts of developments in maize biotechnology on the comparative advantages of developing-country producers; a study of the changing organization of cocoa and rice biotechnology research, carried out as part of a larger research program examining the way in which changes in public/private sector balance associated with economic reforms are affecting the prospects for innovation and enhanced productivity in developing-country agriculture; and, most recently, a six-country study looking at the ways in which biotechnology could contribute to more sustainable crop protection and production approaches.

The overall objective of these studies, Brenner notes, has not been to predict patterns of impacts, but rather to examine the factors stimulating or impeding the development and diffusion of new technologies, and the kinds of policies and/or institutional arrangements that might improve the situation. Results of Development Centre research suggest that the spread of biotechnology applications has been constrained by lack of capabilities in particular scientific disciplines, frequent lack of innovative financing mechanisms, limited attention to marketing and scale-up issues, and (at least in some cases) weaknesses in the regulatory framework at the national level, particularly with regard to intellectual property protection and biosafety procedures. Partly because of the relatively small number of biotechnology-based products currently on the market, impact assessment remains a difficult and uncertain undertaking.

Moreover, she notes, pinpointing the contribution of biotechnology to a given output may be difficult, because new crop varieties may be the result of a combination of biotechnology and more traditional plant breeding and seed production techniques. Efforts to engage in ex-ante assessment should devote greater attention to demand-side issues, rather than simply concentrating on supply-side problems and constraints. There should also be an effort to experiment with alternative approaches; in particular, Brenner argues that the notion of "national systems of innovation" applied in a number of recent studies on industrial innovation may represent a promising approach for analyzing patterns of development and diffusion of agricultural biotechnologies.

The paper by Regina Galhardi takes a somewhat narrower approach to the topic, by assessing the possible trade-related employment impacts of developments in agricultural biotechnology. Although there have been a variety of reports outlining the possible threats to developing countries as a result of the substitution of biotechnology-based products for traditional developing-country commodity exports (see Junne 1992), there are few estimates of the potential magnitudes of impacts on employment.

Taking the cases of coffee and cocoa in Costa Rica, Galhardi develops a set of alternative scenarios for domestic production, export, and employment, based on different assumptions regarding the evolution of world demand. The results of these simulations show the potential for significant direct job losses; as much as 48% for coffee and 27% for cocoa in the more pessimistic scenarios. Even in the more optimistic scenarios, where declining export demand is offset in part by increases in domestic consumption, the country would still experience significant direct job losses in coffee and cocoa production. Negative employment impacts will persist regardless of trends in labour productivity over the study period - and would be much higher for other producers in Central America, which operate at much lower levels of labour productivity than is the case for Costa Rica.

As Galhardi notes, this exercise is by its very nature speculative. It depends on available information on labour coefficients for individual crops and on critical assumptions regarding the future evolution of exports, consumption, and production. The use of national aggregates may gloss over important differences in labour intensity by size of plot, cultivation techniques, and region. Moreover, the analysis deals only with direct - as opposed to net - employment impacts, and does not consider the possibility of offsetting price movements that would alter demand for particular crops. Nonetheless, the paper provides a tentative illustration of the potentially important employment impacts of advances in agricultural biotechnology and, at the same time, underscores the complexity of efforts to assess these impacts.

The final paper in this section, by Michelle Chauvet, presents some of the results obtained from an ongoing effort by a team of Mexican researchers to analyze the socioeconomic impacts of agricultural biotechnologies. The group has completed studies of applications in livestock production and flower-growing, with additional studies under way examining sugarcane and potato production. The studies are of particular interest because they represent one of the few attempts to date to engage in ex-post analysis of the impacts of specific biotechnology applications in developing countries.

In the case of livestock (beef, dairy, and poultry) production, the researchers examined the impacts of a variety of biotechnology-based products: a composite cattle-feed made from agricultural by-products, probiotics for livestock and fodder crops, the synthetic growth hormone somatotropin (rbST), and embryo transplant techniques. In each case, the researchers found, the spread of biotechnology-based products has been relatively slow, despite their potential impacts on costs and/or productivity. In the case of rbST, for example, the hormone has an immediate impact on yields, but also necessitates higher investments that cannot be recouped at prevailing market prices and that are beyond the reach of many small-scale producers.

Similarly, in the case of flower production, high initial investment requirements have slowed the spread of biotechnology applications. Where new techniques have been applied, they have resulted in substantial job creation. High royalty payments to (monopolistic) technology suppliers, however, have restrained the growth of output and have served to maintain downward pressure on the wages of the primarily female labour force engaged in flower production.

Methodological Tools and Approaches

The second group of papers deals with the kinds of methodological approaches that could be applied in future analyses of the impacts of agricultural biotechnology in Latin America. Once again, there is considerable variation among the papers in terms of the kinds of approaches recommended, their potential applications, and their requirements in terms of data-gathering and analysis.

The paper by Otero takes the broadest - and most provocative - approach of the three. He argues that biotechnology could play an important role in the transition to an alternative, more sustainable form of agricultural development in Latin America, but that "the real social forces behind technological and product development respond to different dynamics than the discourse of social and environmental sustainability."

Much of the existing literature assessing the potential impacts of biotechnology, he claims, presents an optimistic scenario open to question on at least three grounds: its assessment of the available scientific and technological resources in individual developing countries, its assumption that agricultural biotechnology will be neutral in terms of its impact on the economic scale of production, and its expectations regarding the possibilities of scientific and technological cooperation among developing countries to develop and diffuse biotechnology-based products.

Otero argues that current economic and institutional trends (notably the oligopolistic structure of input producers in the agrifood sector) are likely to result in a pattern of application of biotechnology that exacerbates rather than reduces social and regional polarization and extends rather than supplants dependence on chemical inputs. In the face of these forces, he argues that the role of sociological analysis is not simply to identify impacts, but also to "identify areas of social organization where some policy and action may be effectively directed." This in turn demands new approaches to impact assessment.

Otero concludes his chapter by outlining two possible approaches. The first of these emphasizes the alternative types of global "commodity chains" through which biotechnology products may be channelled (notably the distinction between producer- and buyer-driven chains); the second emphasizes the importance of studying what Otero refers to as "bottom-up linkages," that is, the extent to which local environmental and social forces are taken into account in technology development. In both cases, the emphasis is not simply on understanding the forces shaping patterns of development and diffusion of agricultural biotechnology, but also on demonstrating opportunities for the expression of different sets of interests and forces, which could alter the trajectory of biotechnology applications, and hence their social and environmental impacts.

Like Otero, Heloisa Burnquist argues that the goal of impact assessment is not simply to outline possible impacts, but also to assist in identifying ways in which negative impacts can be minimized and positive impacts enhanced. Moreover, and again like Otero, she argues that impact assessment should be based upon an understanding of the economic trends and actors shaping patterns of technology development and diffusion. Beyond this, however, the two papers differ markedly in terms of their suggested methodological approaches.

Based on an analysis of current trends in the Brazilian agricultural system, Burnquist argues that, at least in Brazil, impact assessment should be guided by a concern for impacts upon income distribution and household food security. She then proceeds to outline three potential approaches to the study of the impacts of particular agricultural biotechnology applications, each of which presents different advantages and disadvantages.

The first possible approach, a qualitative technology assessment model based on construction of alternative scenarios, has the advantage of being able to respond to a variety of questions not easily amenable to quantitative analysis; its primary disadvantage lies in the subjective nature of many of the conclusions.

A second, more quantitative approach involves the construction of a simple structural model with demand and supply equations for the commodity(ies) under study and simulation of the impact of changes in productivity on prices, demand and supply by crop, and region and type of holding. Such an approach offers a more objective analysis of impacts (although, of course, key variables in the model will be based on more or less subjective assumptions by the researchers), and permits researchers to experiment with different assumptions regarding key variables (adoption rates, for example) as well as with the impacts of possible countervailing policy actions. Its primary draw-back lies in the need for primary data collection and the resultant need to limit analysis to certain products and/or regions.

A third and final approach would be to use an input–output model to simulate intersectoral interactions resulting from the introduction of a particular biotechnology application. Although such a model can be useful in illustrating intersectoral impacts not easily captured in other approaches, its utility is limited by the lack of up-to-date input–output matrices in many countries of the region and by the fact that the model does not account for the impact of changes in relative prices.

The final paper in the section, by Max Colwell, outlines the methodology applied in a recent Canadian study of the impact of recombinant bovine somatotropin (rbST) on the Canadian dairy industry. This study represented one aspect of a broader assessment of the expected impacts of rbST in Canada with other components reviewing animal and human health issues, possible impacts on animal genetics, and a review of U.S. consumer reaction since the introduction of rbST in that country in February 1994. (For an overview of the other aspects of the Canadian rbST assessment, see rbST Task Force (1995).)

Colwell's paper illustrates the need to combine a number of different methodological tools and approaches to adequately assess the impacts of a given biotechnology application - even in a case like this, where the study was concerned with a relatively restricted set of impacts. The Canadian study of impacts on the dairy sector consisted of five components, each of which demanded a different methodology:

· An aggregate analysis of the impact of alternative adoption scenarios on key variables (production, consumption, prices, farm cash receipts, expense and income, and quota values), based on simulations using an econometric model of the Canadian agricultural sector;

· An estimate of the impact of rbST on the Canadian milk supply management system, based on costs of milk production figures taken from farm-level data sets;

· An assessment of the financial impacts on individual milk producers, based on a study of 130 benchmark farms;

· A study of the possible impacts of nonadoption of rbST on Canada–U.S. competitiveness, based on a comparison of the benchmark farms with farms of a similar size in New York state; and

· An assessment of the potential implications for the dairy processing industry, based on qualitative analysis and interviews with key informants.

Integrating Impact Assessment Data into Decision-Making

The third set of papers in the volume deals with the challenge of integrating impact assessment data into decision-making processes. As noted at the outset, impact assessment is not simply an academic exercise but, instead, can provide crucial information to decision-makers at a variety of levels. Just as clearly, however, concerns for the uses (and users) of impact assessment cannot be simply added on as a final step in the analysis but must be integrated into the design and application of impact assessment tools from the outset.

The paper by Catherine Halbrendt and colleagues deals with the possible uses of impact assessment as a tool to improve decision-making within national agricultural research systems. The authors argue that there are three distinct levels of decision-making affecting biotechnology research: individual scientific research projects, national policy and planning decisions, and decisions concerning international collaborative ventures and potential foreign markets. In each case, the kinds of assessment information and analysis needed by decision-makers differ - from studies of consumer demand and "willingness to consume" regarding specific products, to the evaluation of possible impacts of biotechnology products on key national goals (growth, income distribution, employment), to the monitoring of technological trends and market characteristics in other countries.

The authors then proceed to outline a six-step model of the decision-making process for national agricultural research systems and to discuss the kinds of data requirements for effective decision-making at each stage of this model. Finally, the paper reviews some of the activities of the Intermediary Biotechnology Service (IBS) in support of improved decision-making with regard to biotechnology research - from the directory of expertise maintained by the IBS, to a series of policy seminars for decision-makers in developing countries, to a number of commissioned reports dealing (among other topics) with the economic impacts of developments in cocoa biotechnology.

The second paper, by Joyce Byrne, addresses the issue not from the perspective of those concerned with improving research decisions, but rather from the point of view of public sector officials charged with overseeing a regulatory review of biotechnology-based products. Byrne outlines some of the general principles applied by the Canadian government in regulatory decisions regarding biotechnology products and reviews in more detail the process followed in the rbST case.

As with many of the other papers, Byrne underscores the complexity of the issues at stake, and the need for a transparent decision-making process that affords opportunities for participation by all interested parties. Indeed, she argues that in the rbST case, it was precisely the lack of an accepted process for public input into regulatory decisions that contributed to heighten political concern over the issue, elevating the debate beyond the level of dispassionate technical analysis. Byrne argues that there is a need to examine means of ensuring effective public input into regulatory decisions at an early stage and refers in particular to proposals to establish an ongoing "socioeconomic forum" as one promising approach.

The final paper in this volume, by Terry McIntyre, discusses ways of integrating environment and sustainable development concerns into decisions regarding biotechnology applications. Throughout the workshop, participants noted that methodologies and approaches for the study of socioeconomic impacts were much further advanced, and more widely used, than was the case for environmental impacts. This is clearly an area where additional work is required, and McIntyre's paper offers some preliminary suggestions as to how environmental considerations could be integrated into government programs to promote biotechnology-based applications.

McIntyre's paper sets out seven broad points to be considered in the environmentally-sound management of biotechnology:

· Regulatory initiatives providing for prior assessment of the potential implications of biotechnology products before their release;

· Development and application of institutionalized biosafety criteria;

· Ecosystem management approaches;

· Considerations of sustainable development (as opposed to simply environmental impacts);

· Impacts on biodiversity and, specifically, considerations related to the 1992 Biodiversity Convention;

· Provisions for public awareness and public input into decision-making; and,

· Ecological risk assessment.

In each of these areas, McIntyre outlines a series of more specific questions that would ideally form part of a process to integrate an environmental dimension into research and regulatory decisions regarding biotechnology products.

Directions for Future Impact Assessment Studies

[I am grateful to Rodolfo Quintero, Charles Davis, and Bill Lesser for chairing and reporting on the working group discussions. This section is based largely upon their summaries.]

As noted earlier, the concluding day of the workshop was given over to working group discussions regarding the kinds of approaches that could be pursued in future impact assessment studies. As a means of organizing the discussions, two groups were formed, each of which focused their attention on one of the priority areas for biotechnology applications identified in the Mexican priority-setting exercise (Solleiro and Quintero 1993) - on the one hand, bio-pesticides and, on the other hand, veterinary vaccines. (The Mexican project outlined three priority areas for future biotechnology applications in the agrifood sector in Mexico: biopesticides, veterinary vaccines, and treatment of livestock wastes.)

The results of the working group sessions illustrated the complexity of the issues at hand and the diversity of possible approaches to designing a program of impact assessment studies. Of the two groups, the biopesticides group took the broadest approach, focusing on the key variables to be used as assessment criteria and on a series of general suggestions regarding the organization and implementation of any future research program.

On the first point, the group returned to the so-called "5 Es" introduced by Catherine Halbrendt during the previous day's discussions:

· Employment

· Equity (particularly effects on distribution of income and productive assets)

· Environmental impacts

· Economic growth

· "Eating," i.e., food security

The group argued that at least in the case of biopesticides, stress needed to be placed upon social impacts, notably impacts on employment creation and on distributional issues (including possible impacts on land tenure patterns). Not surprisingly, they also argued that there was a need for much greater attention to the assessment of the likely environmental impacts of biotechnology applications; although approaches to the study of economic and social impacts are fairly well developed, we are at a much earlier stage in the development and application of tools for ecological risk assessment.

The group also put forth a series of suggestions on the broader issues of how to organize a research program in this area:

· Impact assessment work should be integrated into a broader process of consultation with key stakeholder groups (government, business, farmers' groups, environmentalists, consumers' advocates, etc.).

· Attempts to assess possible impacts should begin with an assessment of the regulations affecting the use of biopesticides or other biotechnology applications, and should also be geared to suggest possible improvements in regulatory frameworks (for example, the integration of ecological risk assessment).

· Similarly, background research is needed to understand the business and technical conditions of the bio-pesticide industry (including the factors affecting technical change and the nature of competing technologies) to assess the likely patterns of diffusion of new applications.

· Impact assessment work should be coordinated with a broader public awareness campaign, drawing on the kind of stakeholder group mentioned earlier, and should focus on the factors affecting public acceptance of new biotechnology-based applications.

· Much greater attention should be placed on ex-post assessment of the costs, benefits and risks of specific biotechnology innovations; this will demand ongoing monitoring, and should focus on both real and perceived costs and benefits.

· This entire process should be monitored and documented so that the conclusions can be fed into the development of future policy frameworks for the application of biotechnology in the agrifood sector.

The second working group, focusing on veterinary vaccines, concentrated more specifically on methodological issues. The group developed a six-step technology assessment model that could, with modification, be applied to a number of fields of biotechnology applications.

A Possible Six-Step Technology Assessment Model Step 1: Subsector Description

Step 2: Characteristics of Technology in Question

Step 3: Farm-Level Assessment

Step 4: Projected Farm Adoption Rates/Patterns

Step 5: Public/Consumer Response

Step 6: Aggregate Analysis and Policy Options

Step 1: Subsector Description

A rapid overview of particular subsector (e.g., poultry, dairy) focusing on:

· Nature of production (what is produced? where? size distribution of farms?)

· Current patterns of technology use

· Linkages, including structure of input supply

· Regulatory framework

This kind of analysis would be based on existing data sources and would focus on qualitative description of the subsector rather than statistical detail.

Step 2: Characteristics of Technology

A review of what is known about the specific technology under question, focusing on the following issues:

· Expected effects (positive and negative), including potential environmental consequences

· Costs

· Sources and extent of adaptation required for production

· Means of distribution/administration

· Entry barriers - managerial, skill, capital requirements

· Regulatory environment - health, intellectual property, etc.

· Anticipated future rates of technological change

This kind of information should be relatively easy to access in cases of public sector technologies. In the cases of proprietary technologies developed by private sector companies, however, it may be more difficult to gain access to all of the required information.

Step 3: Farm-Level Assessment

Researchers would have to assess the likely impact of the new technology on individual farms (output, input requirements, employment, profitability, etc.). At least three approaches could be followed, alone or in combination:

· Formal sector models. Where formal sector or subsector models exist, these can be a useful means of outlining anticipated impacts. Given data requirements, construction of new models is not feasible, but a certain amount of updating and/or adaptation of models may be required.

· Representative farm analysis. Particularly where a good typology of farm characteristics already exists, this approach can be used to solicit information on likely impacts in a small number of farms; extension workers can often be used as a means of gathering information from individual farmers.

· Qualitative analysis. Especially in cases where the introduction of new technologies is unlikely to produce changes in the overall production system (e.g., product substitutes), qualitative estimates of likely impacts may be helpful.

This stage of analysis is likely to be the most demanding in terms of time and data requirements.

Step 4: Projected Farm Adoption

Once researchers have arrived at a picture of likely impacts on individual farms, the next step is to analyze the probable rate and pattern of adoption of the new technology to arrive at an estimate of the aggregate supply response. For existing technologies already in use elsewhere, this kind of analysis can usually be undertaken using partial adoption models to project adoption trajectories. For new technologies, however, this stage is likely to be more complex and would require estimates of adoption rates by different categories of farms based on available data on costs, capital requirements, access to information/training, policy framework, cultural factors and attitudes of farmers, underlying product demand, etc. Although much of this information will already exist, some additional survey work may be required.

Step 5: Public Response

The fifth step in the analysis would be to examine anticipated patterns of public (consumer) response to the new technology. In cases where we are dealing with existing products (i.e., where only the underlying process has changed), this is relatively straightforward. Even in cases where survey data on public opinions toward biotechnology is not available, national and international regulatory standards can be used as a proxy for public opinion.

In the case of new products, analysis is likely to be more complicated. In addition to secondary sources, some small-scale survey work may be required, either direct public opinion surveys, or interviews with representatives of consumer advocacy groups and/or other experts. Data will be required on food availability, perceived environmental or human health impacts, and levels of trust in the government's regulatory process. In either case, analysis should encompass both domestic and export markets, and thus some information on target export markets will be needed.

Step 6: Aggregation

As a final step effort should be made to arrive at an understanding of aggregate impacts, based on the work undertaken in steps 1–5. In particular, research should attempt to provide estimates of the total production impacts in the regions studied, and distribution of effects across farm types, and between adopters and nonadopters. In addition to impacts on aggregate output and input requirements, this should also focus on potential employment impacts, with particular attention to differential impacts on male and female employment. Research should also analyze the distribution of costs/benefits among broad groups (producers, consumers, suppliers, processors).

Aggregate analysis should also focus on the following:

· Implications for consumers - availability, cost, quality, safety, and dietary implications; and

· Environmental implications.

Finally, existing supply and demand models can usually be employed to provide estimates of how the technology under question will affect producers in other countries and regions, as well as likely impacts for producers in substitute and complementary products.

This kind of impact assessment exercise should also devote some attention to the kinds of policy responses that could be employed to deal with the anticipated impacts - actions to encourage or discourage use of particular technologies, compensation to disadvantaged groups, regulatory reform, R& policies, and consumer education.

Clearly, the demands of this kind of model, in terms of data collection, time, and research skills, are considerable. It has the advantage, however, of integrating a number of different analytical tools (quantitative and qualitative) and of being sufficiently flexible to respond to a range of research concerns. Moreover, the general model can be simplified depending on the precise research questions to be tackled, resulting in a more rapid (but somewhat less detailed) assessment of impacts. For examples of how this kind of a model can be abbreviated, see Love and Lesser (1989) or Miles et al. (1992).

Conclusions and Next Steps

In the end, the workshop made considerable progress in clarifying some of the strategic choices facing any effort to design and implement a program of biotechnology impact analyses: the balance between ex-ante and ex-post analysis*; the variety of levels of analysis and key variables that can be chosen as a focus for analysis; and the trade-offs between alternative assessment methods in terms of the types of questions to which they can respond, and their requirements in terms of time, data, and research skills.

[*On this point, there seemed to be general agreement that more ex-post assessment would be desirable, but the relatively slow pace of introduction of new biotechnology-based products makes this unlikely. Although there are some obvious opportunities for ex-post analysis (e.g., work on rbST in Mexico), much of the emphasis will (and should) be on ex-ante assessments. Nonetheless, ex-ante analysis should also be viewed as an opportunity to collect baseline data for possible future ex-post analysis.]

There was also some progress in outlining promising impact assessment techniques, as reflected in the working group discussions summarized earlier. Clearly, however, there is a need to consolidate knowledge regarding some of the existing approaches and to explore some newer, nontraditional methods that were suggested, but not developed in detail, at the workshop.

Given the relatively incipient state of research in this field, a strong case can be made for methodological pluralism and for experimentation with a variety of approaches either on their own or in combination. As noted, the workshop discussions also highlighted the need for more careful development and application of environmental impact analysis techniques, to match the ongoing work on socioeconomic impact analysis.

More broadly, the workshop underscored the need for a consultative approach to impact assessment to ensure that the broadest possible range of interests is reflected in the analysis. This is essential not simply to improve the quality of analysis but also to ensure a strong constituency of public support for resulting policy decisions. Ensuring a balance between scientific rigour and public participation is by no means an easy task, but it is essential to ensure the effectiveness and credibility of impact assessment exercises.

Finally, and more specifically for IDRC, the discussions during the workshop identified a number of areas in which immediate research could serve to advance the state of knowledge regarding the impacts of biotechnology and to explore alternative assessment methods, for example, further work on rbST in Mexico, or work on the biotechnology applications for cotton and papaya mentioned in the paper by Burnquist. IDRC intends to pursue some of these opportunities immediately and to develop a broader program of impact assessment research as the CamBioTec initative moves forward.

References

· Cohen, J.I. 1994. Biotechnology priorities, planning and policies: A framework for decision-making. Intermediate Biotechnology Service, International Service for National Agricultural Research (ISNAR), The Hague, Netherlands. ISNAR Research Report No. 6.

· Fransman, M. 1991. Biotechnology generation, diffusion and policy: An interpretive survey. United Nations University Institute for New Technologies (UNU/INTECH), Maastricht. UNU/INTECH Working Paper No. 1.

· Junne, G. 1992. The impact of biotechnology on international commodity trade. In DaSilva, E.J.; Ratledge, C.; Sasson, A., ed., Biotechnology: Economic and social aspects - Issues for developing countries. University of Cambridge Press, Cambridge, UK. pp. 165–188.

· Love, J.; Lesser, W. 1989. The potential impact of ice-minus bacteria as a frost protectant in New York tree fruit production. Northeastern Journal of Agricultural and Resource Economics, 18(1), 26–34.

· Miles, H.; Lesser, W.; Sears, P. 1992. The economic implications of bioengineered mastitis control. Journal of Dairy Science, 75(2), 596–605.

· rbST Task Force. 1995. Review of the potential impact of recombinant bovine somatotropin (rbST) in Canada: Executive summary. Report presented to the Minister of Agriculture and Agri-Food Canada, May. Ministry of Agriculture and Agri-Food Canada, Ottawa, ON, Canada.

· Solleiro, J.L.; Quintero, R. 1993. Prioridades de investigación y desarrollo en biotecnología agroalimentaria. Centro para la Innovación Tecnológica, Universidad Nacional Autóñoma de México, Mexico.