Cover Image
close this bookPriorities for Water Resources Allocation (NRI)
close this folderThe wider environment
View the documentPaper 15 World food production: the past, the present and the future
View the documentPaper 16 Climate change and the future of agriculture

Paper 15 World food production: the past, the present and the future

T. W. Tanton and R. F. Stoner
Institute of Irrigation Studies, University of Southampton

Summary: The paper examines why growth in agricultural production is now slowing and whether present-day intervention measures are capable of reversing this trend. It is argued that the present methodology used to identify potential development projects is unsatisfactory, and as a result many projects fail because they do not address the key issues that limit productivity of the farms. It is further argued that if we are to halt the decline we must therefore adopt a more systematic approach to discovering the key bottlenecks in the agricultural sector. This would then allow the most appropriate intervention measures to be identified. An alternative planning framework is suggested for the agricultural sector that could lead to the development of a better method for identifying effective intervention measures.

Past success

In the past 30 years world food production has undergone an unprecedented rate of growth, largely enabling agriculture to meet the needs of the rapidly growing population of most countries outside Africa. This success story is in contrast to the trends in the preceding years, when increases in production could not meet the expanding demand for food.

This success story has resulted from a number of key factors which combined to give an unprecedented increase in production:

· New lands were brought into production by the rapidly expanding population. In all countries this has traditionally been the main way of producing more food to feed a growing population and, in most cases, when land is plentiful this provides an economic solution. Unfortunately, with the increased pressure on land, new farmers are forced on to more marginal land, resulting in a rapid degradation of much of the world's soil resource, our short-term success being partly at the expense of long-term sustainability.

· The introduction of short straw, disease tolerant varieties of wheat and rice, capable of responding well to artificially high levels of fertility.

· The rapid expansion in area of irrigated land, with its potential to produce consistently high yields, has played a significant role in increasing agricultural production during this period; the area of irrigated land now being some 220 million hectares.

· Increased use of agrochemicals.

· Increased mechanization.

Although there are other contributing factors that have been important for increasing the production of specific crops and in solving problems occurring in unique locations, there can be little doubt that the above mainly account for the unprecedented success of agriculture in the last 30 years. The relative importance of each of the above factors differs from country to country depending on local conditions, and where one or more of these variables have proved limiting or inapplicable they have been unable to benefit from the so-called 'green revolution'. For example, sub-Saharan Africa has not been able to benefit significantly since wheat and rice are not the staple food crops, and surface water supplies suitable for irrigation are restricted.

Clearly, despite the overall success story we have had significant failures and we need effective intervention measures to overcome the production constraints in many countries, particularly in Africa. We must also consider if this unprecedented rate of growth in agricultural production can continue.

Where are we now?

A nation's land and water resources are finite and in many countries these resources are becoming fully utilized and, hence, the potential for further extensive expansion of agriculture is limited, with farmers already developing the more marginal lands and utilizing poor quality water for irrigation.

Improved agricultural inputs, the so called 'green revolution' technologies, have enabled farmers to utilize the land more intensively and to obtain better crop yields the technology being particularly important in those countries dependent on wheat and rice. From a slow beginning in the early 60s the technology has been increasingly adopted, and in most countries farmers are now aware of the benefits to be gained from it. During this period nearly all farmers throughout the world have progressively adopted the practices which they find to be beneficial in their own environment, have modified others such as the recommended levels of fertilizer to their own socio-economic circumstances, and rejected others. As a result in many countries the unprecedented rate of growth is now slowing, and in others it has stopped, despite the fact that in many countries yields are still at a fairly modest level.

Pakistan is a good example of a nation that has benefited significantly from all these intervention measures, enabling it to feed its rapidly growing population (growing at 3.1% per year). Unfortunately, this impressive rate of growth has slowed in the past 10 years, with the yield of the major food crops remaining fairly static; increased production resulting from an increase in the irrigated area (Table 1). Nevertheless, yields remain relatively low and if present trends continue Pakistan will move from having a grain surplus to an annual deficit of 23% by the year 2000, and a 30% shortfall by the year 2012. Large deficits are also predicted for other crops'. In the years of rapid expansion growth was largely achieved by a progressive increase in irrigated area and by the gradual introduction of new varieties and the use of fertilizer. Production is now levelling off as water resources become fully utilized and as farmers both plant improved varieties and apply levels of fertilizer that they find attractive. Although there is some scope for further production from irrigating more land, the availability of water is likely to pose a major obstacle to further large-scale development of land. As with many countries, if they are to increase agricultural production to meet the needs of a growing population they must achieve more intensive production on existing farmland and use water resources more wisely. The potential to double or more these existing low yields exists, and the technical know-how to achieve this is available and relatively simple, unfortunately as in many other cases the technology proves to be inapplicable because of constraints in the farming system. The result is that we do not know what intervention measures might bring about a significant increase in production, nor do we have a rational approach to identify what the true bottlenecks are and how we might overcome them.

The example is fairly typical of many countries which achieved rapid growth in agricultural production in the past but which are predicted to have severe difficulties in meeting the growing needs of their population unless they find appropriate intervention measures for their farming systems. Unfortunately, all too often many governments try to overcome the problem by developing investment strategies to increase production which are founded on political will or unreliable analysis, with little or no analysis of the sector as a whole. A clear case of this is to be seen in Egypt where scarce financial and water resources are being used to irrigate impoverished sand to grow wheat in the desert.

Table 1 Wheat and rice production in Pakistan

Can investment strategies remove the bottle-necks in agricultural production of the world?

The key intervention measures outlined above provided a simple broad-brush approach to agricultural development with new widely applicable technologies being readily transferred from country to country. Although the approach has had great success it is clear that it has been far more effective in some countries than others, depending on the relevance of the technology to the local environment. The development sector appears to feel that because of the general success of this broad-brush approach a similar strategy can be followed for the implementation of a wide range of possible intervention measures. The result has been a tendency to take the success of a project in one country as a basis to justify the implementation of projects of a similar nature across the globe, with agencies tending to follow each other in investing in projects of a similar nature at any one particular time, the so-called 'bandwagon effect'. Listed below are some of the areas that have been favoured for investment at various times during the past 20 years, they come and go, often having little long term impact:

- Fertilizer subsidies
- Integrated rural development projects
- Extension
- T and V extension systems
- Irrigation
- Dryland agriculture
- Small-scale projects
- Large-scale projects
- Rural credit banks
- On-farm water management
- Financial restructuring of the economy
- Adaptive research Sector-based education.

and more recently:

- The environment
- Women
- Good government
- Project-based training.

There appear to be a number of ways new bandwagons are initiated. They may be based on a very successful project in a specific location or result from a combined political will for projects in a specific area; they can arise from a desperation to try to do something new that might work. There is nothing wrong with any of these and it is well worth trying new ideas on an experimental basis. The problem occurs when the need to spend large amounts of money with the least administrative costs results in the experiments becoming large projects which are rapidly replicated as the bandwagon begins to roll.

On mentioning these reservations to a banker in the aid sector the reassuring reply was that there is no need to worry because all projects undergo thorough economic analysis and have to be 'bankable', giving an internal rate of return of at least 12%. Although the discounted cash flow techniques used for economic assessments provide an accepted procedure for comparing the viability of projects, the results can only be justified if the assumptions on which they are based can be justified. Unfortunately, all too often this is not the case, and on close examination the economic viability of a project is often pivotal on a number of professional judgments which the evaluation team has had to make. Given that in most developing countries we do not know with any level of certainty what the underlying bottlenecks in the agricultural sector are, nor the magnitude of the effects, many of our judgments must be considered open to question. Further, in the development business we are all being asked to formulate and evaluate projects in an ever decreasing amount of time, giving little opportunity for us to understand truly the complexity of the agricultural sector of a country for which we are formulating investment polices. We therefore put forward for discussion the idea that the level of uninformed judgment that is required to formulate many agricultural development initiatives largely invalidates much of our economic analysis, with the result that in many countries, despite vast investment in the agricultural sector, production continues to stagnate. We all start out as optimists, not realizing that optimists and pessimists are the same thing except that the latter are better informed.

The international hotels of all developing countries are full of 'experts' in the development industry, all trying to improve the performance of the agricultural sector, but with none of us fully understanding the key issues that are causing the bottlenecks in the system. We readily fall back on the well-used phrase "in my opinion" to make up for lack of facts; the result is that in all too many of our decisions are too subjective. As a result many development initiatives are based on preconceived ideas with little understanding of the multifaceted and complex systems which interact to control production on the farms. Hence, projects end up not addressing the real constraints that exist in a given agricultural system, simple because these constraints have not been recognised.

On asking a wide range of staff connected with a large irrigation system, from the minister to the smallest farmer, "why are yields so low and where should money be invested to increase productivity?", every person identified a different cause and a different solution depending on our personal perspective of the project. In reality we had all identified many symptoms but not necessarily the underlying cause of the problems.

This week's conference has brought together a large group of informed expertise from within engineering and natural resources with a brief to combine the expertise of the two disciplines to enable us to explore where irrigation fits into the agricultural sector and where it should be going in the light of such important issues as global warming and desertification. Depending on our outlook we expound the virtues of different development strategies such as the need to protect the environment, the benefits of small schemes, of large schemes, the role of NGOs and of the cost/benefit of irrigation development as compared with investment in dryland agriculture. The reality is that all have their appropriate time and place; the skill is determining when and where it is.

It is argued therefore that if agriculture is to meet the needs of the world in the next 30 years we will need to develop a more systematic approach to identifying the key constraints in the agricultural systems of the world to enable us to target our aid much more effectively. Such an approach would supersede the need for debates like i' Are large irrigation projects better than small? Should irrigation be favoured at the expense of dryland agriculture?".

A look at possible future investment strategies

Value for money is essential when aid money is targeted at the poorest people on earth. Unfortunately, in recent years it has become more and more difficult to target aid effectively in the agricultural sector. All too frequently money has been spent on projects that have not removed the key constraints in the agricultural sector and it is clear that if aid is to be more cost effective in the future it must be targeted at the key bottlenecks.

There appear to be two main potential investment strategies which could focus investment into effective projects.

· To identify a number of key constraints which are known to limit production in many countries, but to which we do not readily have solutions, and concentrate a large proportion of the aid money into developing solutions. This is a high-risk strategy and projects would need to have the potential for very high returns. Such a strategy might result from a feasibility study to evaluate the technical possibility of genetically engineering cereals to enable them to live with symbiotic bacteria for nitrogen fixation.

· If such a programme were estimated to have only a 50% chance of success it could be argued that the potential benefits could justify spending a large proportion of the world's total aid budget on it. Such high-risk high-gain strategies are unlikely to prove attractive to donors for a number of reasons but they do have many attractions, not least the size of the potential benefits.

· If the managers of a car plant decide that they want to increase production they do not just make investment in the different areas of the production process and hope that it will remove the bottlenecks that will allow increased production. They carry out detailed analysis of the production system to enable them to identify the constraints and the investment needs that will allow them to increase production. Further production constraints are progressively identified until a level of investment is identified that will allow them to reach production targets, i.e. until marginal costs exceed marginal benefits.

Within the agricultural sector of most countries such an approach is not considered. Indeed in many countries the agricultural production system is often directed from within two separate ministries, irrigation and agriculture. There is often both an irrigation sector five-year plan and an unrelated agricultural sector plan. Detailed examination of a number of such plans from several countries indicates that there is normally no rational basis for the plan but that they are made up of a mixture of the aspirations of the civil servants and the wishes of the politicians. In reality at the end of the planning period they prove to have been little more than a work of fiction. These poorly thought out plans are used to identify the investment opportunities for development agencies and form the basis for requests for aid and/or soft loans.

We clearly need to move towards a more rational planning process in a similar systematic fashion to that used in manufacturing industry. The World Bank Œries out regular sector reviews, as do many aid agencies, but these fall far short of a full analysis of a country's agricultural sector in which the key bottlenecks are identified and quantified.

Many agricultural researchers have recognized that if they are to produce research results that meet the farmers' needs they must adopt a systematic approach to identify on-farm constraints, thus enabling them to identify their research priorities. The analytical techniques that they have developed are proving powerful tools and it is logical that such a systematic approach should be extended to enable us to identify the true bottlenecks in a nations agricultural sector; thus enabling investment to be targeted accurately.

This approach implies a rigorous, quantitative and systematic effort, first to understand the farming operation, and then proceeding to define and develop an appropriate investment strategy. Farming systems need to be analysed in their entirety to establish which of the multitude of variables impinging on farm activities are the main causes of low productivity. This appears to be the only way to identify which constraints need to be removed, if it is possible to remove them, and how.

For this analysis the farm population needs to be disaggregated by such factors as agroclimatic zone, size, tenure systems and the availability of water. The effect of social issues need to be quantified. The economics of the farming practices of the disaggregated farms need to be fully understood. Analysis is largely possible through statistical and survey techniques developed by farming systems researchers and others. Such farm-level models, however, have severe limitations if used in isolation and the analytical approach has to be extended to examine the constraints that exist in domains outside the farm. It is at this level that the approach is likely to need further refinement.

Such a systematic approach to analyse the agricultural sector of a country would bring together the skills of all the disciplines involved in effective agriculture development to produce a comprehensive reference document on the state of a nation's agriculture. Such documents would adopt a systematic approach to analyse the agricultural sector, starting from identifying issues that affect the production of individual crops and management of a farm's resources, to studies of the potential to utilise further the nation's resources in the political and economic climate of a country. It is expected that such a document would be updated annually and act as reference work to provide analysed data for the further development of agricultural sector plans; both for the effective identification of projects and also to improve project formulation.

Although such a systematic approach is needed in many countries it is suggested that the methodology should be fully developed and tested in one country. If successful it would then act as a blueprint for use in other appropriate environments.


It was pointed out that a conclusion made at a conference held at Wye College two years ago was that targeting aid at the emergent commercial farmers was more likely to produce substantial benefits than attempting to target the poorest but ODA policy has continued to overlook this conclusion. It was argued that it is not a question of choosing between the top down scientifically driven strategy or the site-specific bottom-up approach analysing constraints on farmers but both should be pursued. In discussion over possibilities of further dramatic increase in yields in South Asia it was argued that good use of water was the key. It was stated that we are on the verge of breaking major scientific barriers to higher yield, such as the photosynthetic barrier, but crop yields of up to 6 tonnes/ha, though easily achievable, were rarely being attained in the developing world, well below the 10-20 tonnes/ha set by the photosynthetic limit. It was added that the major issue is not so much the upper limits to achievable yields but the yield gap between average farmers and poor farmers. Under-yielding is a problem common to many industries besides agriculture. It was argued that there are common socio-political problems which donors attempt to address through influencing policy. The concept of 'yield per unit of water evaporated' was suggested as a useful criterion. It was recognised that improvements in water-use efficiency could be made but also that excess water passes into the groundwater and is re-used. Despite the importance of fish in diets (up to 90% of animal protein consumed in Bangladesh) many of the benefits of increased rice production are lost through decline in fish catches due to irrigation practices. It was suggested that proponents of irrigation needed to take a wider perspective but also pointed out that experience in India shows that aquaculture and rice irrigation can go hand in hand.