Mechanisms for assessing priorities in animal production research.

Abstract

Proc. Aust. Soc. Anim. Prod. Vol. 17 MECHANISMS FOR ASSESSING PRIORITIES IN ANIMAL PRODUCTION RESEARCH H. LLOYD DAVIES* INTRODUCTION 103 The topic of this contract is of interest to everyone in this audience whether he or she be producer, strategic research worker, applied research worker, extension officer or educator. There is a hierachy in assessing priorities in research for animal production. The first assessment is a national one to decide whether our research effort should be directed toward the primary industries or should our research efforts be into the manufacturing industries, the service industries or other industries, e.g. the communication industry? If we assume that agriculture is given a high national priority there then has to be some method of determining priority areas in agricultural research - should it be soil conservation, cropping, animal production, processing, marketing, etc. If we accept that animal production research is given a high priority within agricultural research we wish to know what are the mechanisms for determining the priorities in research in animal production. Members of the panel will discuss some of the mechanisms used by their organisations to assess priority allocation to an animal research activity, e.g. pasture production, disease control, genetics, nutrition, reproduction research etc. There would be agreement that scientists must have what Ronayne (1984) refers to as 'professional autonomy', namely that universal criteria are used to evaluate contributions and that the scientific results are shared within the community. If we accept 'operational autonomy' for the research workers, the function of Departments of Agriculture, the Board of the C.S.I.R.0. and the Rural Funding Councils must be to decide on which problems to study and the resources allocated to them. There can also be little doubt that in having decided priority areas, funds should be allocated, if possible, to 'guaranteed achievers' namely those who have an established record of completing a research project and publishing the results. As well as selecting a competent researcher those who allocate funds must have confidence that the 'selected achiever' will work on projects that would have some hope of industrially useful application. The Panel represents a broad spectrum of people who have had to make had decisions about the allocation of priorities and have had to:be answerable to their constituencies whether it be the'producers who are funding the research, the manufacturers, the consumers or thescientific community'in general. An absentee from the, Panel is a University representative. In the Universities, the internal research funds are so small that it is extremely difficult to pursue animal production research.without seeking funds from ~outside organisations and,, therefore, University workers are constrained by the priorities se`t by the Rural Research Trust Funds. Do these constraints on the Universities jeopardise the supply of trained researchers for future animal production requirements or should a portion of research funds be, allocated for training? 'This is one of the many difficult questions to which there are no easy answers. *Department of Wool Science, School of Fibre Science and Technology, University of New South Wales, P.O. Box 1, Kensington,* N.S.W. 2033 104 Proc. Aust. Soc, Anim. Prod. Vol. 17 PRIORITY SETTING FOR WOOL RESEARCH A. J. FARNWORTH* The allocation of research funds may be viewed as similar to any business investment decision and subject to similar analyses/recommendations. Research 'investment' analysis is more difficult because of the uncertain nature of potential benefits and because, in many cases, the output is knowledge rather than a specific project to market. In the wool industry, this uncertainty is increased by the large diversity of potential research investments ranging from the fundamental level to joint venture/commercialisation at virtually all stages of the wool 'pipeline' from farm production to final product development and manufacture. In the early days of-administering wool research funds, priorities were set on a totally subjective basis at the project level, e.g. is it a worthwhile project aimed at a known industry problem?. do the researchers appear capable of solving the problem? ; does project A appear to be more worthwhile than project B? More sophisticated and objective approaches have evolved over the years which, to varying degrees, facilitate informed discussion and evaluation. However, the ultimate decisions necessarily continue to have a large subjective element. PRIORITIES WITHIN RESEARCH AREAS The Wool Research and Development Council bases its priorities and fund allocations within research areas on recommendations from advisory committees for production, harvesting, measurement, distribution, textile and economic research. The committees comprise individuals selected on the basis of personal skills and expertise as users of research and/or providors of research and/or administrators/managers of research. Although none of the committees has adopted a formal or explicit investment analysis framework in determining research priorities, they all address in at least a qualitative and sometimes quantitative manner particular factors which are relevant to and consistent with such an' approach. The 'user' members can provide input on the magnitude and immediacy of the problem, the potential benefit per farm or industry sector, probable rate of adoption of a successful innovation. The 'provider' members can assess the probability of success dnd cost of the research, the 'management' members advise on the capacity and ability of the proposed researcher, scope.for co-ordination . with other research or funding organisation etc. The Production Research Advisory Committee, supplements these deliberations with comprehensive reviewti of specific fields of research together with comparative refereeing and ranking of projects. PRIORITIES BETWEEN RESEARCH AREAS ' A much more difficult problem to,which the Wool Research and Development Council is giving considerable attention is the determination of priorities 'and optimum allocation of funds between broad research areas. *Wool Research and Development Council, 530 Little Collins Street, Melbourne, 3000 Proc. Aust. Soc. Anim. Prod. Vol. 17 105 In recent years attempts have been made to adapt methods used in business investment decisions to increase the objectivity of research fund allocation procedures. This requires quantification in monetary terms and ranking systems for programme benefits (size of 'industry' affected, size of productivity improvement, rate of adoption, leakage of results, distribution of gains, market price impact, etc.) and costs (cost of research, development and adoption, foregone opportunities, risk and uncertainty factors, etc.) These efforts have so far failed, partly because no existing methodology was suitable for direct translation of wool research,but more particularly because of the difficulty in quantifying many of the essential factors. A simpler approach has been in use for several years - namely 'funding at the margin'. Indicative budgets were provided for each research area, with a substantial overall contingency set aside to fund projects considered worthy of support and which, comparatively, offered greatest industry benefit but could not be funded within the indicative budgets. This comparison of marginal projects from several research areas was expected to transfer funds over time from lower to higher priority research areas. In practice, there has been little impact on the ratios of funding between the areas of production, harvesting, measurement, distribution, textile and economic research, primarily because most of the worthwhile projects could be funded within the indicative budgets, leaving relatively few for comparative evaluation. Additionally, even where the opportunity to compare projects arose, the unavailability of objective, comparative information led to decisions being centred primarily on project details rather than relative benefits to industry. The Wool Research and Development Council is making a renewed attack on this difficult and complex issue. A project has already been commissioned `to try and establish a system for determining the distribution of benefits from any innovation to various industry sectors - e.g. how much of' the benefit from research on an increase in processing productivity accrues to growers as distinct from processors, consumers etc. The Council also intends calling for competitive tenders for a consultancy to develop a general decision framework for determining the magnitude of benefits to woolgrowers of wool research and development * activities in terms of cost reductionand/or demand increase, effects on supply and demand interactions,. price implications, etc. MEAT AND LIVESTOCK RESEARCH AND DEVELOPMENT P. McINNES* ; INTRODUCTION The Australian Meat and Livestock Research and Development Corporation (AMLRDC) was established by a Commonwealth Act in 1985 and is the only R. and D.. . Corporation for an Australian rural commodity. The functions of the Corporation are research and development with the object of may be of use for the purpose of improving storage, transport or marketing of meat or directed to meat and livestock acquiring and applying knowledge that any aspect of production, processing, livestock. In developing priorities for such research and development, the Corporation *Australian Meat and Livestock Research and Development Corporation, Aetna Life Tower, Cnr. Elizabeth and Bathurst Streets, Sydney, N.S.W., 2000 106 Proc. Aust. Soc. Anim. Al)od. Vol. 17 consults with organisations such as : Cattle Council of Australia; Sheepmeat Council of Australia; Australian Meat Exporters' Federal Council; Meat and Allied Trades Federation of Australia; The Council of Australian Public Abattoir Authorities; Australian Livestock Exporters' *Association. These organisations are acknowledged as contributors of the levies which fund the activities of the AMLRDC. ESTABLISHMENT OF PRIORITIES Issues and constraints, and methods of solving problems in the meat and livestock industry were sought from both industry and research organisations within a few months of the establishment of the AMLRDC. The 120 replies were examined and analysed technically and commercially by the Directors of AMLRDC. The AMLRDC Board has some experience in industry, and commercial and government R. & D., and was selected by industry representatives as a sound team to appropriate funds to increase efficiency and effectiveness in the meat industry. The replies formed the basis of the Corporation's Research and Development Plan which defines the principal objectives of the Corporation during the period from 1986 to 1990. The broad strategies which are being pursued to achieve these objectives have been published and distributed to industry and R. & D. organisations. Five principal objectives were identified in this Plan. These are : 'acquiring knowledge about markets; improving the efficiency of converting livestock to meat; improving the efficiency of livestock production; improving product range, quality and marketability; improved management of the R. & D. programmme. Subsequent funding by the AMLRDC has followed this Plan. In 1986 some sections of the industry approached the AMLRDC and discussed the subject of the most appropriate level of funding for R. & D. for the meat and livestock industry. The Corporation resolved that a study be commissioned via the Cattle Council of Australia to seek the optimal level of such funding. A Steering Committee was established which included, in addition to the Cattle Council, representation from sheepmeat interests, meat exporters and the AMLRDC. The subsequent report 'Review of Research and Development the Australian Cattle and Sheepmeat Industry' , prepared by ACIL Ltd. recommends an immediate increase in total funding of R. & industry of $lOM and justifies such expenditures by examples of R.'& D. programmes. R. & D. PRIORITIES AND FUNDING . To set priorities is an exercise reflecting industry problems. Funding . R. & D. to solve these problems is restrained by the quality of applications, and the restriction on the amount of finance available. The Corporation has considered in 1986 and 1987 project submissions addressing these problems from ' traditional organisations such as CSIRO; Universities and Departments of Agriculture and Primary Industries. In the Corporation's view, many submissions lack firm objectives or commitments to solving problems. r The approach of the Corporation in addres,sing priorities is increasingly being directed to commercial organisations where a firm contractual agreement cati be achieved. Expenditure in Australia Pty. D. for the cost/benefit Proc. Aust. Soc. Anim. Prod. Vol. 17 SETTING RESEARCH PRIORITIES FOR THE ANIMAL INDUSTRIES - A CSIRO PERSPECTIVE A. D. DONALD* THE PAST 107 CSIRO's research priorities have tended to evolve in an informal, 'bottomup' way within a historical framework of research divisions. Funding decisions by the Rural Industry Research funds and informal interactions between CSIRO scientists, livestock producers and other scientists have had important effects on research directions. In 1978, CSIRO tried a more formal process for setting agricultural research priorities involving internal consultations between Divisional Chiefs, invited external input and a study which aimed at providing an economic framework for priority setting (Johnston and Girdlestone 1983). The process of research resource allocation in CSIRO, and in the old RIRF s, conformed largely to the precedence model (Anderson and Parton 1983). in which upward and downward adjustments are made from the previous year's funding. This maintains continuity in particular fields, but tends to perpetuate past emphases, with the risk that fields of declining productivity will continue to be supported. Anderson and Parton (1983) 'questioned whether research resource allocation based on 'the precedence model is sufficiently sensitive to promote a set of research activities that adequately reflects the pressing problems of agriculture'. This is a reasonable question when the imperative is the solution of immediate problems. However, for longer-term strategic research which may lead to completely new technologies, it is arguable whether any system which did not depend on imaginative ideas developed by research scientists, and supported by continuity of funding over a series of years, would have delivered benefits to industry more efficiently. THE PRESENT The restructuring by the Government of the RIRF s has been a significant change with emphasis on strong producer input, accountability, increased commercial returns from research, and the development of strategic plans. These plans show changes from the past in their strong market orientation, and their intention to devote resources to more efficient management of R. & D.'programmes. I CSIRO introduced strategic planning in 1984, with the first Divisional and Institute plans being developed for 1985/86. While the Organisation cannotyet be said to be'strongly driven bystrategic planning, its introduction led to the appointment of Divisional Advisory Committees composed predominantly in the case of the animal research divisions, of livestock producers. These committees have already had a significant impact on priority setting at research programme level within CSIRO. CSIRO's strategy document, 'Shaping the Future: A Strategy for CSIRO 1985090', has two strategic objectives relevant to priority setting, namely: to develop more systematic procedures for identifying growth areas and assessing the balance of research across economicsectors; and to introduce more systematic evaluation of research. For the first, a number of criteria have been defined, including the potential of an industry to generate wealth and employment and contribute to exports; the promise of major scientific advances and the likelihood that they will be taken up; 'and the availability of resources and skills. \ . *Institute of Animal and Food Sciences, CSIRO, PO Box 225, Dickson, ACT 2602 108 Proc. Aust. Soc. Anim. Prod. Vol. 17 . Although unweighted and largely non-quantitative, these criteria provide a framework which could be useful for assessing competing claims for research resources between the different commodity areas of the animal industries. For the second, a Research Evaluation Committee, comprising senior research managers, commissioned a study by McKinsey and Co which developed a framework for ex ante evaluation of commercially applicable research proposals. The framework ' involves identifying the business system to which the research relates and assessing the potential commercial benefits, including the ability to capture that benefit,for Australia. Analysis of the sequence of activities by which products. are produced and delivered to markets (the business system), and the dominant economic forces which affect them leads to an understanding of the key points of economic leverage and indications of the sorts of applications which will be most valuable. This evaluation framework with some modifications is presently being extended within CSIRO to test is applicability to various research areas. THE FUTURE Ex ante evaluations, to determine research priorities are intrinsically subjective and can impose substantial costs if complex, time-consuming approaches are used. Nevertheless, there is scope for improved methodologies in setting priorities for the animal industries at different levels. Analysis of the business systems for wool, meat, dairy products, and so on, is undoubtedly a useful first step, although it is not really new. The strategic plans of the AMLRDC and'the WRDC contain substantial elements of that kind of analysis, and differences between the wool and meat business systems have led to different emphases. Analysis at this level is largely economic and can be undertaken at various levels of sophistication. 'For industries of such economic importance as wool and meat, and for which substantial amounts of data are obtainable, there may well be value in pursuing a rigorous econometric model ,similar that of Davis et al. (1987) for assessing priorities. It is important . that the subjective scientific judgements for such models are provided by ,knowl,edgeable scientists. The vexed question of priorities for long-term strategic research can to some extent be guided by a business system analysis indicating which areas are most likely to benefit from new technology. However, new technology often provides unexpected enhancements in productivity. Such possibilities, which cannot be quantified at the outset, will always warrant taking some funding risks. I These approaches can provide a new systematic framework for assessing the relative priority between different fields', but within fields there needs to be : more emphasis on integrating mult-disciplinary technical and scientific judgements with user perspectives. An excellent structured approach is the Modified Delphi . technique developed by the AWC for its review of pasture,research, described by Wheeler (1986), and since also used for the AWC review of genetics research. Successful involvement of end users in the priority setting process, advocated by Healy and Love (1987), will be a major advance and wil,l contribute substati- ,I tially to the more rapid adoption of new technology. Proc. Aust. Soc. Anim. Prod. Vol. 17 RESEARCH PRIORITIES - A BUREAUCRAT'S VIEW JOHN C. RADCLIFFE* 109 Research, per se, brings no benefits. It is the application of research results that leads to benefits. The identification of research priorities involves an estimate of the probability of benefits being obtained from the proposed research. Scientific excellence is no longer the primary criterion for management of research. Research in the animal industries must be oriented to producing results which can be applied to the technology of animal production and marketing systems in order to increase their efficiency and allow producers to maintain or improve their relative positions against their competitors. It generally follows that the primary goal or goals of reseach must ultimately reflect monetary benefits. Increasingly, efficiency of resource use rather than yield maximisation is the primary objective. There has been a traditional should derive from ideas generated priorities are first determined by specific tasks which scientists are a widely held view among scientists tifically demeaning and stultifying conflict as to whether research priorities by scientists ('bottom up') or whether senior managers and are translated into required to execute ('top down'). There is that the latter form of management is sciento the researcher (Lindner 1976). A combination of these approaches can be fruitfully adopted and has been incorporated into the managerial arrangements established under the Rural Industries Research Act 1985. Rural Industry Research Councils are required to set broad objectives in their 5 year plans which are circulated to research organisations ('top down') , yet research proposals in response to these objectives primarily derive from scientists in a 'bottom up' process and form the basis of the majority of funding decisions. Although Councils have the right to independently commission research ('top down') such programmes appear to be a minor part of most Councils' research portfolios. However, some Councils are developing proposals to commission a greater proportion of their research programmes. Criteria to be taken into account in setting research priorities include the cost of the research in relation to the likely benefits; the probability of an innovation being achieved;, the duration for and extent of its likelladoption if successful; who appropriates or shares the benefits; and'the extent of any disbenefits (for ex&nple on the environment). The skills and resources of those seeking to undertake the work and their previous 'track record' must also be considered, along with proposals advanced for the conmlunication,of the results.. to encourage adoption of thenew technology. The setting of research'priorities is widely recognised as subjective. The propensity of final plenary sessions of research conferences to list all possible topics as priority areas so that no members of the peer group are disadvantaged is well known. Therefor,e, numerous attempts have been made to quan- I, tify selection criteria by developjng scoring systems. Most, however, are structureless and arbitrary. Although many have been proposed, few have been tested and even fewer have`beenused for any length of time (Greig 1981). They may have some value in encouraging a degree of rigour and consistency into the priority setting process provided the values obtained are not treated as absolute determinants. Benefit:cost analyses of the likely gains from the proposed' *Department of ,Agriculture, Adelaide, S.A., '5000 110 . Proc. Aust. Soc. Anim. Prod. Vol. 17 research can also be useful, not as absolute values in themselves but rather to give a guide, by the way the process is approached, as to whether the researcher has any understanding of the industry for which the research is being proposed. All too frequently, researchers appear to have a very limited understanding of the commercial realities of the industries they are seeking to help. A mathematical tool which can be used for ranking subjective opinions is the Delphi technique. This is a structured procedure for forecasting, and attempts to remove or control problems of inadequate representation, deference to authority and the sociological factors of group interaction such as 'bandwaggon' effects, Rothschild (1972) highlighted the need to identify the clients of research. Their interests have often been represented through complex advisory committee structures. These are expensive to operate. The members may have no effective responsibility for the ultimate research priority decisions and are likely to become disillusioned. It is noteworthy that CSIRO has abandoned the use of bodies such as the State Advisory Committees. Clients must. have effective particiption in the research priority-setting process. This usually will mean some control over the budget. There may also be opportunities for client groups 'to be represented in staff selection processes. Client representatives now play an effective role in the project selection and accountability processes of the Rural Industry Research Councils. Similar processes are extending to other fund sources. In South Australia, much of the funding for pig research is allocated by a representative producer committee set up under the Swine Compensation Act, with the result that there is a very close proprietary and effective relationship between pig producers and 'their' researchers. At a time when most government resources are being reduced in real terms, research funds as a result of recent changes to research legislation and the greater commitment of rural producers. to sponsor research, are more readily available than funds for advisory services. This may well lead to an imbalance between research activities and extension services'in the years ahead. Greater priority will need to,be given to communicating the results of successful research and encouraging its adoption. It is possible for research teams which have had little change or stimulation to lose their inno,vation and creativity. It is therefore desirable to periodically review the objectives of research groups. Experience in South Australia where a number of major research reviews were undertaken with client representation in the early 1980s showed that it can 'take some while for-a research group to digest the results of such an experience and for there to be an impact on its productivity. Some sectors of the scientific:community in Australia appear to have been reviewed with such frequency and;given such mixed signals that any benefits have been largely lost, priorities have become confused and the scientists disillusioned. -Priority setting in animal production must ultimately be a 'co-operative and iterative process between client producer groups and researchers working in response to broad objective&identified- by the producers themselves. QUANTITATIVE TECHNIQUES FOR ASSISTING RESEARCH POLICY PLANNERS \ . J. L. BLACK* Scientific`r'esearch is both expensive in terms of resources 'required and *CSIRO, Division of Animal Production, P-0. Box 239, Blacktown, N.S.W., 2148. . Proc. Aust. Soc. Anim. Prod. Vol. 17 1 1 1* slow in terms of the time taken to solve problems in sufficient detail for the results to be adopted by society. Currently, within Australia, there are insufficient resources to allow scientists to work only on subjects of interest and it is essential to set priorities for research. There is an hierarchy for making decisions on research policy. Priorities must first be set at the international or national level where choices are made on a broad scale between such areas as the environment, economic efficiency, defence, human health and welfare or uncommitted basic research. Then within each broad area, for example, economic efficiency, decisions must be made about the relative benefits from research into agriculture compared with manufacturing industry, transport and others. Within agriculture, there are decisions about the value of additional research into animals compared with crops,'pastures, soil conservation, irrigation and so forth. Then within animals, decisions must be made about the species and, within each species, about the form of production, the physiological process and, finally, the specific research form of production. Decisions made higher in the hierarchy are frequently politically motivated, whereas those made at the lower levels tend to be based more on fact and logic. Quantitative techniques can greatly assist the efficiency of decision making at all levels. The main quantitative techniques for aiding the establishment of research priorities are the survey or epidemiological study involving physical measurement and the integration of knowledge using computer simulation models. Both methods provide quantitative information that allows decisions to be based largely on current knowledge and they reduce the chance of wrong decisions resulting from either subjective intuition or political bias. Surveys and epidemiological studies are most suited to upper levels of the research decision making process, whereas the simulation techniques can be used at all levels. Once the broad area for research has been established, two criteria should) . be satisfied before finally selecting a research project: (i) alternative ways of meeting the goals should be investigated and the one most likely to result in the fastest progress should be selected provided that the resources and expertise of the researchers are adequate; and'(ii) all current knowledge in the area and associated areas should be taken into account before the precise aims of the research, the hypothesis to be tested and the experimental protocol are finalised. Thig will ensure that new knowledge required to improve understanding of the area isobtained and will help to avoid unnecessary duplication of previous research through ignorance. It is important to consider information at a more fundamental level than that of the nominal research area because an understanding of the underlying mechanisms may explain apparent anomalies in research findings., Computer simulation models provide an excellent way of helping satisfy these criteria. priorities I In 1968, the Club of Rome (Meadows et al. 1972) used a computer simulation model to examine the limits to growth on this planet. Predictions from the model indicated that the world system would collapse early inthe 21st Century. The initial reason for collapse was predicted to be depletion of natural: resources . leading to a shortage of energy, a decline in food production and excessive 'pollution, Substantial technological advances such as unlimited energy supply through nuclear power, a reduction in birth rate-to 25% of 1970 level, doubling food production and a recycling 75% of non-renewable resources were predicted to delay the time of collapse from one to several decades. The world system was .stabii'ized only when population and capital investment were controlled simultaneously at 1975 and 1990 levels, respectively, along with substantial technological achievements. The anticipated population for 1975 was 3.5-billion; it is already 5 billion. The model suggests that highest priorities should be given to, equalizing birth and de&h rates and to reducing,the gro+wth in * National-international research 112 Proc. Aust. Soc. Anim. Prod. Vol. 17 industrialization. An increase in agricultural production alone would delay collapse by only a few decades but it was predicted to be an essential component for ultimate population stability. This result may sound encouraging for animal scientists but there are adverse consequences of increased animal availability. The atmospheric concentration of methane, which is partially derived from ruminants, is increasing rapidly and is many times more effective than carbon dioxide in increasing the earth's temperature and computer predictions indicate that this could have disastrous effects on sea level a