Long-Term Projections of Global Food Requirements: Why Were We Wrong? ()
1. Introduction: Global Agriculture in the 21st Century and Long-Term Projections
The run-up of global food prices since 2004 that reversed a three-decade old pattern of declining or stable prices is causing concerns in developed and developing countries alike, as well as in international organizations such as the World Bank and the United Nations’ Food and Agriculture Organization (FAO). Higher prices for food and energy—the two principal components of economic wellbeing—are putting at risk the accessibility, availability, and affordability of these two critical resources for millions of the world’s most vulnerable people. Some of the gains made over the last decade or two in reducing the number of people living in poverty worldwide have been lost due to the sharp increases in food and energy prices. Even though the FAO’s global food price index at the end of 2012 has receded modestly from its peak in 2010- 11, currently, it is still 100% higher than it was in 2000 [1]. As a consequence, since the 2007-08 civil unrest in some 30 food-short developing countries [2], the issue of “food security” is becoming as important as the issue of “energy security” in the international public policy debate.
Increasing geo-political tensions among the leading economic powers who are trying to lock-in their resource supplies, along with mounting scientific evidence of climate change due to the emissions of greenhouse gases, is making it clear to political leaders and policy-makers that transformational changes in the global energy, transport, manufacturing, and agricultural systems will be required if living standards are to increase over the next 40 years, when the world’s population is projected to reach more than 9bn people [3]. Consequently, there is a critical need for long-term projections of these resources—because of the long lead-times to develop additional food and energy capacities—if rising geo-political stress and domestic civil unrest are to be contained.
Before turning to the subject of this article, it is important to provide the context for writing it. In September 1981, a symposium sponsored by the Norwegian Nobel Institute was convened to “explore the long-term perspectives of world demographic and economic growth with particular emphasis on international differentials in demographic and economic characteristics, on resources and supplies, and on implications for emerging patterns of cooperation and conflict” [4].
This article reviews long-term global projections made more than 30 years ago of four agricultural sectors: livestock products (including meat, eggs, and milk products); oil crops (which include oilseeds, peanuts, and soybean products, among other crops); grains (which include maize (corn), rice, wheat, barley, and oats, among others); and root crops (which include potatoes, sweet potatoes, cassava, and yams, but not sugar crops), with the observed data from FAO’s detailed Food Balance Sheets [5].
2. Long-Term Studies on Agriculture, Food, and Nutrition
Two studies published within a year of each other that focused on long-term trends provide the facts to support the view that without a successful global agricultural sector that incorporates production, transportation, distribution, and changing consumption patterns, improved material well-being in the future—characterized by higher life expectancy, improved nutrition, higher incomes, and lower poverty rates—will not be possible.
Robert Fogel [6] used statistical data—with particular emphasis on Europe and America—to highlight the remarkable extension of life expectancy in the 20th century and the decline in mortality and morbidity rates which he attributed largely to increasing (and nutritionally-improved) food supplies that provided the higher energy levels needed for a growing and increasingly healthy work force to fend off a long list of, often fatal, infectious diseases. In turn, these improvements in human nutrition contributed to economic growth and development and technological change, and vice versa. Without this synergy, it is doubtful that mortality and morbidity rates would have declined sufficiently to achieve the remarkable average life expectancy levels that we are enjoying today.
Giovanni Federico [7] reviewed the success of agriculture over the last two centuries. Federico argued that the world agricultural system has successfully fed an ever-growing population with an increasing variety of products at falling prices, and, simultaneously over the 200 hundred years since Malthus’s hypothesis [8], it did this while releasing a growing number of workers from agricultural work to the rest of the economy. According to Federico, the world was able to feed 6.5bn (in 2005), up from 1bn in 1800, because of steady progress in the “primary forces”—scientific advance, technological and institutional change, and globalization.
However, Federico questioned whether the still growing population in the developing countries can be provided with the nutritional standards of the developed countries that reflects a much higher proportion of meat and dairy products that are much more land-intensive than a diet comprised primarily of cereals.
Another important issue that Federico raised is the front burner issue of the “sustainability of modern agriculture”, that is, how do we mitigate the environmental damage caused by the excessive use of fertilizers and pesticides, diminishing bio-diversity, increasing scarcities of water and land, and increasing levels of greenhouse gas emissions, with the continuing improvement in the nutritional value of food in particular, and material wellbeing in general, in Asia and Africa in the 21st century?
The US Department of Agriculture (USDA) publishes 10-year projections every year of commodities, trade, and prices based on projected US and world economic and population growth, agricultural policies, and other assumptions that affect the demand and supply conditions in the global agricultural sectors. The United Nations’ Rome-based Food and Agriculture Organization recently published long-term projections of global food and nutrition levels for 2050 [9].
Two other organizations that develop long-term projections of global food and agriculture requirements are the Washington-based International Food Policy Research Institute (IFPRI) that is supported by the Consultative Group on International Agricultural Research (CGIAR), and the International Institute for Applied Systems Analysis (IIASA), in Laxenburg, Austria, that conducts policy-oriented research on problems that are both global and inter-disciplinary in nature. Their recent work on long-term projections of global agricultural requirements to 2050 appears in the FAO report, “Looking Ahead in World Food and Agriculture: Perspectives to 2050” [9].
In addition, a number of recent policy papers by government and academic experts [10,11] and review articles in the international press have appeared [3,12,13] on the challenges and opportunities for farming and food to the middle of the 21st century.
3. The United Nations World Input-Output Model
In the early 1970s with the recognition of the increasingly adverse environmental effects caused by worldwide industrialization and the first major oil crisis in the decade, the United Nations voiced its concerns regarding the growing gap in per capita incomes between the less-developed countries of the world and the highly industrialized ones.
As a result, in 1973 the United Nations commissioned the construction of a general-purpose model of the world economy. In hindsight, writing in mid-2012, this modeling effort should still be recognized as an intellectual “tour de force”. As far as the resource sectors were concerned, the model—which was developed by Professor Wassily Leontief, who was awarded the 1973 Nobel Memorial Prize in Economic Science—tracked three fossil-fuels, six metallic minerals, and four agricultural resources. Other agriculturaland food-related variables included in the model were irrigation investment, land development, cultivated land area, calories and proteins per day, and fish catch. The model also included 30 manufacturing and service sectors, as well as eight types of major pollutants, and five pollution abatement activities [14].
Economic activity was regionalized—the countries of the world were originally aggregated into 15 regional blocks—but were unified by export and import flows of goods and services, capital flows, aid transfers, and “cross border” payments of interest on borrowed capital. An upgrade of the model around 1980 separated the two countries—Canada and the United States—included in the “North America” region, making a total of sixteen regions.
Once assembled, the model was designed to provide quantitative projections of regional and global resource requirements, pollution levels, cumulative resource use, and required inter-regional financial flows, etc., under varying assumptions regarding future income growth in the developed and developing countries with a view towards narrowing the income gap between the two groups of countries from 12:1 (in 1970) to 7:1 (by the year 2000). Curiously, according to the observed per capita GDP for the year 2000 used in the tables in this article, the income gap between the two groups of countries declined from 12:1 in 1970 to 6.6:1 in 2000.
4. Projected and Observed Population and Income Growth: The Main Drivers of Global Food Requirements
This section presents the principal long-term determinants of food consumption—population and income growth. The population projections made in the mid-1970s for the year 2000 by the United Nations Population Division [15] were insignificantly different from the observed global rate for 2000 (Table 1).
This table presents the projected and observed annualized regional population, Gross Domestic Product (GDP), and GDP per capita growth rates for the projection interval. (Please see Tables A1 and A2 in the Appendix for the country and region aggregation schemes). The lower observed population growth rates for most of the regions (with the exception of Africa, Asia, one European region, and the United States) as compared to the projections are consistent with long-term demographic trends. The demographic changes in two of the European regions WEH (western Europe, high income) and EEM (eastern Europe, medium income) are the result of the re-unification of Germany in the early 1990s and the migration of many eastern-Europeans to western Europe, in part, as a result of the economic stress in the eastern countries in the wake of the collapse of the Soviet Empire in the late 1980s. As for the US, the higher observed population growth rate compared with the projected rate for the projection interval is most likely the result of increased immigration during the last two decades. The observed growth rate for (non-Japan) Asia exceeded the projected rate by less than 10%, and the only substantial “error” appears to be Africa (AF), where the observed rate exceeded the projected rate by 0.5% per year.
Observed world GDP growth was less than the rate projected by the World Model, despite the formidable growth rates achieved by China (ASC), and India and other Asian countries (ASL), from 1980 to the terminal year, 2000. The observed lower growth rates in GDP relative to the projected rates over this interval in the middle-eastern oil-producing countries (OIL), Central and South America and the Caribbean (LA), South Africa (SAF), the Soviet Union (SU), and Eastern-Europe (EEM), and the lower observed growth in Japan and Western Europe (WEH), can be attributed to an array of well-known political, social, economic and financial problems that these regions confronted during the 1980- 2000 interval. These problems also explain, in part, the lower observed world GDP growth rate relative to the projected rate despite the successes in Asia.
The last two columns of Table 1 present projected and observed annual growth in GDP per capita. For economists, per capita GDP represents an imperfect, though adequate, metric to measure relative “living standards” and their growth (or decline) over time.
In addition to population and GDP growth, regional and global food requirements are also impacted by other important variables such as urbanization rates, which along with higher living standards, result in changing diets such as a movement away from root crops to grains, and finally, to increased livestock products, characterized by more meat, eggs and cheese, in daily diets.
Despite not having access to the documentation regarding the technical relationships governing these regional changes in food diets from decade to decade over the projection interval when preparing this article, it is my understanding that because the World Model tracked both the changing regional urbanization rates and per capita GDP levels from decade to decade, the projections of the regional food requirements were, in part, determined by these factors that were modified region-byregion and decade-by-decade.
5. Global Agriculture, 1970-2000: The Projections and the Observed Data
5.1. Production and Domestic Supply
This section provides an overview of the projections and observed data for the 30-year projection interval. Please see Appendix B1 for some caveats about the data, and Appendix B2 for a comment about “consumption” in the World Model. Tables 2 and 3, respectively, present the projected and observed annual growth rates in production and domestic supply (and, in the case of the observed data, food consumption) for the four agricultural sectors.
Table 1. Projected and observed growth in population, gross domestic product, and gross domestic product per capita, 1970-2000 (annual percentage change).
At the global level the projected annual growth rates are remarkably close to the observed rates, save oil crops, where the observed growth rate is almost four times higher than the projected rate. I comment on this discrepancy in Section 6, below. It would be reasonable to expect greater divergence at the regional level, not only between the projected and observed production values, but also between the values for regional domestic supply because of import, export (and, in the case of the observed data, inventory adjustment) activities. This is indeed the case. Section 6 examines the differences at the regional level between the projected and the observed growth rates employing the central explanatory variables— population, GDP, and GDP per capita growth—as the “primary suspects”. At the global level, the observed growth rates of livestock products, oil crops and grains all exceeded the rate of growth in population and per capita GDP, with only the rate of growth in production of root crops lagging population and per capita GDP growth.
5.2. Regional (or Country) Dependence on Food Imports
“Food security” has become an issue for countries who rely on food imports to bridge their shortfall between domestic production and food consumption. The World Model also tracked regional import and export activities. Therefore, it would be of interest to examine the changing position of these regions over the 30-year period with respect to their dependence on imports to supplement their domestic food production. It is important to remind readers that the degree of import dependency of a region—or more specifically of a country—on energy or food is only one factor among others that comprise the more complex terms “energy security” or “food security”. Some of these other factors include the sources of those imports, the availability of substitutes, the ease or difficulty of increasing efficiency in their use, etc.
The changing rates of projected and observed dependency on agricultural imports are presented in Table 4. While there are relatively large differences between the projected and observed data regarding the degree, i.e., the percent, of import dependency in 2000, generally speaking there is broad agreement between the projected and observed data regarding the regions that are most dependent on imports to supplement their domestic production.
5.3. Per Capita “Consumption” of Livestock Products and Grains: The Projections and the Observed Values in 1970 and 2000
Table 5 presents the projected and observed annual per capita domestic supply (in kilograms) of two of the four