The modernization of agriculture on the African subcontinent has for the most part been disappointing, and for a long period of time. The hypothesis advanced here is that the failure of agricultural modernization in Sub-Saharan Africa is rooted in (1) an inappropriate bias toward resource-saving effects in the technology promoted as the means to modernize the agriculture in the region, (2) a rather misguided and excessively orthodox bias when assessing the benefits and costs of subsidies used to promote the use of fertilizer and other modern inputs in the region, and a failure to develop appropriate institutional arrangements that are essential to the modernization of the sector, especially in an open economy.
I. RESOURCE ENDOWMENTS AFFECT ADOPTION OF NEW PRODUCTION TECHNOLOGY
A prominent feature of the discussion of agricultural development for Sub- Saharan Africa is the perceived need to promote a Green Revolution in the region similar to the one that occurred in Asia. The emphasis is on producing and adopting yield-enhancing crop varieties and the technological package that would support them. Other elements of the package include the use of modern fertilizers, pesticides, and irrigation.
The World Bank and other international development agencies such as the Sasakawa Global 2000 program have also sought to develop extension services that could promote the use of that technological package. Some attention has also been given to changes in economic policy that would provide stronger incentives for farmers to adopt the new technology and would integrate the domestic economy more effectively into the international economy.
However, the promotion of this technological package and the other instruments of agricultural development have, with a few exceptions, been notably unsuccessful in the region. Food production in sub-Saharan Africa has not kept up with the growth in population, with the result that food production per capita has, overall, declined. There is limited progress in alleviating poverty in the region, and the economies for the most part have stagnated in aggregate terms.
It is hypothesized here that the failure of agricultural modernization in the region is due in part to a failure to promote appropriate kinds of technology at the farm level. This comes from a failure to appreciate the significance of the underlying resource endowment as a factor influencing the adoption of new technology.
There are two ways of thinking about the decisions pertinent to the introduction of new production technology in agriculture. [For the importance of the modernization of agriculture in promoting economic development, see Schuh (2002).] One way is to recognize, rather simplisticly, the high social rates of return to investments in agricultural research in the past (Alston et al., 2000) and to concentrate on making such investments as the basis of economic growth and development. Most of those studies have been of an ex post nature and have been largely oblivious to underlying resource endowments as they affect the adoption of any new technology. The social rates of return to the creation and introduction of new agricultural technology tend to be high, ranging from 25% to 35% on the low end to over 100% on the high end (Alston et al., 2000). Since most developing countries can borrow money for investing in agricultural research at quite low rates of interest (5% and less in recent years) and on very soft terms, this seems like an obviously good investment.
A second, more operational or instrumental way of thinking about the process is to conceive of the new production technology as alleviating the constraints that impede increases in agricultural output, with an emphasis on which resources in the endowment are particularly important for affecting the adoption of the new technology. Hayami and Ruttan (1970, 1985) pioneered in this way of thinking about the process. Their analysis showed that if labor scarcity is the main constraint to increasing agricultural output, the need is to alleviate that constraint by concentrating on labor-productivity-enhancing technology, such as mechanization. If land is the relatively scarce resource, the new production technology should be land-productivity-enhancing, such as fertilizers. This perspective raises important issues of science and technology policy. In particular, it raises the issue of evaluating the resourcesaving effects of any new production technology, together with a more acute ex ante perspective in the allocation of resources to agricultural research. Hayami and Ruttan also posit a positive (in contrast to the above normative) perspective to this issue. They hypothesize that the process of innovation is responsive to relative resource endowments, and thus that these endowments can be used to make predictions about the pace and resource saving dimensions of the process of technological change. These propositions are formalized in a model of “induced technological innovation” [see Hayami and Ruttan (1985)]. They go further, however, and focus on the importance of having the right institutional arrangements if the resource-saving effects of the new technology are to be focused on an efficient growth path in the resource-resource dimension. Ruttan and Hayami (1984) have extended the induced-innovation perspective by advancing the hypothesis that the institutional arrangements themselves are also responsive to the underlying resource endowments.
The perspectives developed by Hayami and Ruttan can be used for both positive and normative analysis. First, they can be used to explain the resource-saving direction that a new production technology takes as it is adopted in the economy, and to explain the direction that institutional innovations evolve in the economy as economic development proceeds.
Alternatively, the model can be used in a normative manner, as a guide to policy in allocating resources to research and institutional development. From the second perspective, both the research process and the process of institutional change need to be guided by an understanding of relative resource scarcity if development policy is to be efficient and succeed.
Hayami and Ruttan showed in their 1985 book that because early in Japan’s agricultural modernization process the economy was land-scarce and labor-abundant, what was needed for agricultural development was a set of biological innovations consisting of improved varieties and a package of modern inputs, including fertilizers, that would raise land productivity. The United States, on the other hand, early in its development process had a resource endowment that was labor-scarce and land-abundant. This induced a process of mechanization in U.S. agriculture that raised labor productivity.
Both processes proved to be efficient and successful, with the result that these two countries led the process of agricultural modernization around the world, even though the respective resource-saving effects were quite different in the two countries. Even more impressive was the fact that in a later period of their respective histories, the relative resource scarcities of the two countries reversed themselves, and the respective innovative processes also reversed themselves. In the case of the United States, the frontier was eventually closed and land began to become increasingly scarce in a relative sense. This induced a process of biological innovation in agriculture, and the productivity of land became an objective. In Japan, a similar reversal occurred. As industrialization proceeded, labor became increasingly scarce.
The result was a process of mechanization in Japanese agriculture. This leads us to the situation in sub-Saharan Africa where, unfortunately, all too often the thrust of policy-making in the region, especially in science and technology policy, has been to try to replicate there the successful “Green Revolution” of Asia. The emphasis has been on introducing improved cereal varieties and the use of commercial fertilizers. The result has been for the most part a failure, and for an obvious reason—including the one described in the next section. The situation is more like that in the United States and Latin America in their earlier days when there was relative labor scarcity in the economy. Uma Lele (1975) made this point some years ago. Unfortunately, the data to test this hypothesis are limited. Wage data and data on the price of land, which give evidence on relative resource scarcity, although accessible for Japan and the United States are not available for Africa. Given this limitation, Lele made a somewhat different argument with more confidence. She called attention to the low productivity of labor in sub-Saharan Africa, due in large part to poor health and nutrition. One can infer from this that productivity enhancing innovations should be focused on raising the productivity of labor rather than on land, as has been the case.
One of the interesting things about the region is that it has not benefited from the animal-driven mechanization that was such an important phase in other now-modernized agricultures (Pingali et al., 1987) More recently, however, herbicides such as Round-Up have been introduced. Such herbicides make it possible for a farmer to cultivate as many as five hectares of land instead of just one. This represents a significant increase in labor productivity and a significant increase in agricultural output for the region should it be possible to spread the use of this labor-saving innovation throughout the region.
If this labor-scarcity hypothesis is correct, and certainly more empirical research is needed on this issue, the dominant innovation process in sub-Saharan Africa should be either some of the many variants of mechanization, or the use of herbicides to raise productivity of labor. It is not that improved varieties and the use of fertilizers and associated modern inputs cannot raise the productivity of labor. The underlying production functions are obviously not completely separable. However, when the productivity of labor is so low, it seems clear that a shift in emphasis is needed to reduce or eliminate that constraint as a barrier to increasing agricultural output.
It should be noted that the adoption of the fertilizer/high-yielding variety package in practice increases the demand for labor. If there is a constraint on the supply of labor, this can inhibit the adoption of the new package. In a study of such a technical package in Plan Puebla in Mexico some years ago, it was found that the availability of off-farm employment was a determining factor in the adoption of the new technology. Families in which the head of the family was working off-farm did not adopt the technology; those in which the head of the family did not work off the farm did adopt it. Interestingly, the average family incomes of the two groups of families, both adopters and non adopters, were approximately the same (Villa Issa, 1976).
The rapid spread of AIDS in sub-Saharan Africa is giving added impetus to this problem. Projections are that in some countries in the region, the population may actually decline in the future. The impact on the labor force will be even greater, given AIDS’ pattern of mortality. It is especially morbid and in turn debilitating in its effects on the productivity of labor.
Unfortunately, these issues of science and technology policy are sorely neglected in current decision-making, especially consideration of relative resource endowments and the relative scarcity of the main inputs in agriculture.
II. SHOULD THE USE OF COMMERCIAL FERTILIZERS BE SUBSIDIZED?
The problems created by the failure to recognize the importance of adapting the design of technological packages to the endowment (or relative scarcity) of resources in sub-Saharan Africa has been exacerbated by strong pressure from the World Bank and other international development agencies against subsidies for the use of commercial fertilizers. These proscriptions have become part of the general policy of international development agencies to eliminate the use of subsidies and other government interventions. This perspective came to the fore during the economic crisis of the 1980s, when debt burdens became excessive and finding the means to service those debts became important. The international development community turned against import-substituting development policies and shifted toward an export-promoting policy stance.
Efficiency in resource use became a policy imperative, and pressures against any policy distortions grew in importance.
Promoting resource efficiency in the domestic economies of sub-Saharan Africa and integrating them into the international economy were obviously important for an efficient growth policy. However, the prevailing policy proscriptions have ignored some important reasons for intervening in the economy and became too restrictive against the use of fertilizer subsidies in particular. The rigid approach taken by the international development agencies has ignored that there can be externalities that make fertilizer subsidies a rational way of promoting the use of such inputs.
The analytical justification for a bias against subsidizing fertilizers is that resources in the domestic economy need to be used most efficiently if maximum sustainable economic growth is to occur. The argument is that tradable products or commodities should be priced at their border price equivalents (that is, they should have domestic prices equivalent to the prices in the international markets as converted through the value of the domestic currency in foreign exchange markets). Providing subsidies for the use of fertilizers would cause their domestic prices to be less than their border price equivalents and thus would lead to “excessive” use of fertilizer. This in turn would place excessive demands on the market for foreign exchange and thus in most cases would bid resources away from alternative uses that presumably could contribute more to economic growth. Such subsidies also tend to make it difficult to balance budgets in the domestic economy and thus make it more difficult to have an effective macroeconomic policy.
This neoclassic economic argument against the use of subsidies is correct in principle, but I ignores certain externalities that are important elements in the economic environment that producers face when making their production decisions. In particular, the logic of the above case ignores important conditions in the economic environment of most agricultural producers, and especially those in sub-Saharan Africa. These economic conditions constitute important externalities that must be taken into account when deciding fertilizer and subsidy policy. In particular, a case can be made for subsidizing the use of fertilizer, given the following conditions: (1) efficient financial intermediaries for the agricultural sector are absent, so farmers are unable to obtain the optimal amount of credit for their operations; (2) investments in roads and other transportation infrastructure have been inadequate (less than optimal); (3) there is high complementarity between fertilizers and improved seeds, and eventually between fertilizers and irrigation water and other modern inputs; and (4) external pecuniary economies are associated with agricultural modernization, which means that the benefits of such modernization would be widely distributed in the economy, and in favor of the poor.
Fertilizer subsidies were fairly high when recent efforts to modernize agriculture began in sub-Sahara Africa over a decade ago. These subsidies were rooted in at least two of the traditional reasons given for providing such subsidies. The first was a learning-by-doing argument, based on the belief that subsidies provided over a relatively short period of time would induce farmers to try the new input. In the process, they would learn how to use it and become persuaded of its benefits. This suggests that the subsidies could eventually be
The second traditional reason for providing subsidies for the use of fertilizers was an attempt to offset the risk and uncertainty in adopting a new innovation such as improved varieties and fertilizers. Poor farmers are justifiably risk-averse, so subsidies lower the financial stakes for farmers trying new varieties and fertilizer. This is related to the learning-by-doing argument but different in that subsidies would not only help to offset the risk associated with the new input but would make its use more profitable. It could be argued in this case also that subsidies could eventually be phased out.
Eventually, the World Bank and other international development agencies turned against the use of such subsidies, however. The attempt to phase out the use of such subsidies was part of a general attempt to phase out the use of import substituting industrialization policies and to move developing economies toward more efficient use of their domestic resources. A related issue was reducing the recurring budget deficits associated with these subsidies, and the need for institutional reforms to collect more domestic taxes to pay for the subsidies. The deficits created recurring problems in managing domestic inflation.