Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
1
Scope and method
2 Choice and application of economic theory
Introduction
Objectives
of the study
Sampling
procedure
Approach
to the study
Outline
of the book
Until the middle of the twentieth century the history of Africa was not one of a continent acting, but of one acted upon. It was only with the attaining of national sovereignty that the constituent countries of the continent began to achieve some independence of action. Yet independence is not easily won, it is not effortlessly retained; even now, Africans are frequently required to reserve certain areas of decision-making to foreigners, or to relinquish other areas over which they have previously gained control.
Such an area is that of economic policy-making. Typically, the newly sovereign African country is able to choose its own economic policies, and apply them to the best of its ability. Nevertheless, the environment within which the policies have been applied has not been conducive to success: dissent from within and shocks from without have put African economies under great stress. Moreover, the high expectations of the liberated populace have raised the standards which that policy must attain if it is to be successful.
It is not surprising, therefore, that many African countries have been unable to meet the aspirations of their people. It is also not surprising that, in their disappointment, they have had to turn to those with wealth for help.
When help is sought by one or two, the response can take into account the particular characteristics of both grantor and recipient: each programme of assistance is specific to the occasion. When, however, many countries appear on the international scene seeking help, the wealthy - wealthy countries, wealthy institutions - naturally establish rules and procedures governing their response. They systematize or regularize the provision of assistance. Moreover, if the number of appeals places pressure on the administrative capacity of the wealthy to respond, they tend to create new agencies which specialize in the allocation and management of assistance: the IMF and the World Bank are today the two primary agencies. Finally, if the number of appeals is rising sharply, the grantors, in fear of still further commitments and in their superiority, begin to impose conditions upon the help they provide.
The help they do provide can be, and usually is, substantial. Grants, loans and other financial support, and technical assistance from the primary agencies, from the wealthy countries that support them, and from commercial and philanthropic organizations are great attractions for developing countries. But along with financial help come the conditions multi-facetted, nearly uniform and primarily economic in nature, imposed upon the poor African countries by the agencies of the wealthy countries, designed both to raise the productive capacity of the recipients and to achieve other, possibly associated, ideological and political goals. Together, the provision of funds and the imposition of conditions are commonly called Structural Adjustment Programmes.
If the enunciated objective of Structural Adjustment Programmes is to help make the recipients more productive, the enunciated objective of this study can be posed as the question: do Structural Adjustment Programmes help to make a few of the institutions of the recipients more productive? Not all the institutions, for the study is not of sufficient scope to address such a monumental question, but only those institutions which attempt to advance the recipient country's science and technology.
Why focus on the advance of science and technology? The answer is clear, if not entirely lacking in controversy: the productivity of a country's population depends more than anything else in the long run on that country's ability to achieve technical progress. In the short run, to be sure, improvements may be secured through increases in the efficiency with which existing scarce resources are applied, through reallocations of the scarce resources to different sectors of the economy and different types of organizations, and through increased availability of imports financed by foreign loans; but these are once-and-for-all advances. After the short-run improvements have been secured, subsequent progress depends greatly upon the country's developing and absorbing new techniques, upon its making and exploiting advances in science and technology. In the long run it is advances in science and technology that are crucial for economic success.
Hence our concern with science and technology, and with the effects of Structural Adjustment Programmes on their advance. But just as we shall be able to study only a sample of those institutions that are assigned the task of advancing science and technology, so we can only consider a few countries. Since the majority of developing countries have adopted, or are adopting, Structural Adjustment Programmes, there is a large universe to select from. A lack of readily available material on the achievements of institutions engaged in furthering science and technology has prevented us from carrying out the sort of statistical analysis on the macro-economic effects of Structural Adjustment Programmes attempted by several authors (World Bank, 1988a and 1990; Harrigan and Mosley, 1991; Faini et al. 1991; and Stewart, 1991). We therefore limit ourselves to a few countries, choosing Ghana in West Africa, and Kenya, Tanzania and Uganda in East Africa. Like most countries in Sub-Saharan Africa these four are quite poor, with few reserves to cushion themselves from external shocks and internal dislocations. The majority of their populations is engaged in agriculture; their manufacturing is erected on a fragile base; their governments' revenues are scanty and expenditures over-committed; and a large fraction of their overseas earnings are consumed in servicing foreign debt. Adjustment to the IMF/World Bank's satisfaction would be expected to be very difficult, and to take a very long time, years or even decades, rather than months. Like other grave economic matters capital investment, education, health, the birthrate - the phenomena on which we are focusing are not altered on short notice. Nor can our observations, extending over at most a decade, provide other than a guess at what will be the lasting effects of Structural Adjustment Programmes on the pursuit of science and technology.
Definition of terms
Science and technology are imprecise terms; yet if we are to measure the effects of Structural Adjustment Programmes upon them, and their effects, in turn, upon the economies under adjustment, we must consider the terms' meaning.
Figure 1.1 may be of some use in setting limits to the terms 'science and technology'. The figure displays the chief scientific and technological institutions within a country, comprised of those which provide scientific and technical education; those independent institutions which carry out R&D; and producers (firms, government ministries, para-statals, plantations, etc.), a few of which conduct R&D, as commonly defined, and many of which also assimilate and improve upon the techniques they employ. The main flows between these institutions - flows of individuals and flows of information (or knowledge and know-how, which tend, in Sub-Saharan Africa to flow with individuals) - are indicated by the arrows. Like others who write about science and technology (e.g. Forje, 1989: 18) we shall consider all the institutions and their links, human and intellectual, i.e. all the stocks and flows in Figure 1.1, to constitute science and technology. We shall also include, to the extent possible, the mores encountered, i.e. those attitudes, inducements, rules and procedures that set the environment within which science and technology are advanced.
To draw a figure, and to say that our subject is all it contains, is not to make the task any easier nor the measures any more precise. Everyone who studies science and technology recognizes that the terms are not only imprecise but also incomprehensible. Relevant outputs from scientific and technical education and R&D institutions are unquantifiable; those from the educational establishments, to whose graduates numbers can be attached, do not reveal their standards of attainment nor their future potential. Where numbers can be ascribed to the scientific and technical education, and R&D institutions they are of the nature of inputs - so many scientists employed within and R&D laboratory or so many engineers and technicians working on improving a company's manufacturing process. These data give no indication of whether or not the individuals are working together harmoniously, whether or not the results of their efforts will be applied expeditiously. Data of this last sort, qualitative material on the effectiveness of institutions' expenditures on science and technology, emerge only after thorough investigation.
Figure 1.1 Flow diagram of the model of structural adjustment
We have already mentioned that our sample of countries numbers four, from among the larger number comprising Sub-Saharan Africa. For these four countries we shall try to obtain as much data as possible on the overall commitment to science and technology. To place these aggregative measures in some perspective we shall relate them to other more familiar macro-economic data.
Completeness of measure is beyond our means for the institutions indicated in Figure 1.1: we did not have the resources to investigate more than a few R&D organizations nor more than a few firms which, in addition to producing their goods and services, improve upon their techniques. Our sample is thus quite small. The reader should not be too disappointed, however, for the universe of institutions is not so much larger. In Forje's 'Directory' of major research establishments in Africa (Forje, 1989: 267-81) 14 are listed for Ghana, of which we investigated five plus two more not listed; 16 for Kenya, of which we covered two, plus two more not listed; seven for Tanzania, of which we covered three, plus two more not listed; and seven also for Uganda, of which we covered one (with multiple establishments), plus one more not listed.
In total, for the four countries, 44 R&D establishments are listed; we covered 11, plus seven others, in depth. Adding the seven extra establishments to the 44 in Forje's 'Directory' brings the total of 51, and our sample to 18, which is one-third of the universe. Such a small sample does not warrant a sophisticated statistical analysis, so we shall use the quantitative material derived from the case studies as illustrative, attributing to it little more credence than the qualitative material simultaneously gathered. But to write of sophisticated analysis is to broach the topic of methodology to which we come next.
There is thought to be an ideal way to approach a study similar to this one. In brief, one draws upon an established body of theory, which yields, as logical deductions, hypotheses. Some hypotheses may be obvious, some surprising, but all should be relevant to the issues arising in the study. One then collects data with which to confront the hypotheses. Provided that hypotheses have not been falsified by the data, their implications can be drawn; if they fail to pass the tests, the hypotheses must be rejected, the theory reformulated, and data gathering and testing renewed. These five stages formulating theory, deriving hypotheses, collecting data, testing hypotheses against data, revising, and determining implications - may have to be undertaken again and again, before one can be satisfied with the outcome.
In the social sciences the ideal is never attained. Theories are never sufficiently precise: often, two or more incompatible bodies of theory purport to cover the same issues. Even if there is only one generally accepted body of theory, it may not yield enough interesting hypotheses. When the issues to be studied are complex, as they usually are, the theory may be equally complex, so complex as to make it impossible to deduce hypotheses logically, i.e. to determine the general behaviour of the system the theory encapsulates.
Succeeding difficulties that arise in attaining the ideal are a lack of conformity between the elements of the theory (variables, parameters) and the units of observation. The data may not provide the exact measures of the hypothetical terms or, alternatively, the hypotheses may not contain variables which rank large in the data. Even if the hypothetical terms and the categories of data match perfectly, data are seldom available in sufficient number and precision to permit the application of discriminating tests. Finally, the time and resources devoted to carrying out a study are usually so limited as to prevent more than one circuit, or perhaps one-and-a-half circuits of the five-stage scheme.
How close does this study of the effects of Structural Adjustment Programmes on the pursuit of science and technology come to the ideal? We shall attempt to answer this methodological question in terms of the five-stage scheme already mentioned. First, as regards the theory, there is none that perfectly suits our purpose. The phenomena we are dealing with Structural Adjustment, as defined by the IMF and World Bank, and the pursuit of science and technology - are respectively, macro-economic and micro-economic (and macro-political). Theory that contains both macro - and micro-economic variables is rare. In the formulation of macroeconomic theory, the behaviour of individual economic agents is usually neglected or, if it is included, is embedded in a highly simplified system (usually specified to be perfectly competitive) from which a representative agent is detached. In the formulation of micro-economic theory, phenomena external to the agents or to some collection of agents, say an industry, are not accommodated explicitly. In conventional microeconomic theory, for example, structural adjustment would be an exogenous factor, i.e. a set of elements not represented in the theory itself. In between the realms of macro-and micro-economic theory there are a few formulations upon which we can draw, although even in these few there are some unwelcome assumptions.
But we will address these matters more fully in the next chapter, which will be devoted to theory, where we will also observe how numerous and provocative are the hypotheses that can be deduced. Moving on to the third of the five stages of the typical study we can ask now to what extent are the data we can collect likely to be similar in their definitions to the variables in the theory?
At the level of the economy, data, and their counterparts in theories, are often closely connected, if not entirely congruent. Aggregate statistics like GNP, investment, exports and imports, government revenue and expenditures, etc. provide reasonable estimates of the same variables appearing in theory. Prices, e.g. wage and profit rates, interest rates, exchange rates, can diverge, in practice and in theory, both as regards their definitions and their coverage. A few of the theoretical terms used to set standards or support judgements, such as welfare and utility, do not have statistical counterparts.
If there are some difficulties at the overall level in matching theoretical variables with their empirical counterparts, these are as nothing compared to the difficulties in matching at a micro-economic level. These variables which rate so large in micro-economic theory outputs, inputs, prices, costs, profits - are almost impossible to measure at the level of the producer. What is the output of an R&D laboratory? What is the likelihood that one can discover the costs of manufacture of a small, privately owned company? What are any firm's profits? And those assumptions which underlie the theory - e.g. profit-maximizing entrepreneurs directing institutions staffed with knowledgeable, rational and dedicated workers carrying out orders with supreme efficiency; costless and instantaneous shifting of resources from one activity to another in response solely to market signals etc. - can they be brought within the scope of an enquiry? Can data of any kind measure the mobility of resources to inducements that may never have been offered before? Yet, as we shall see almost all theory, the micro-aspects of the IMF/World Bank theory of Structural Adjustment for example, depends for its solution on just these sorts of assumptions. If all economic agents are presumed to exhibit these attributes, the assumptions may be acceptable, in the interest of obtaining solutions to the theory. But if some agents, say those operating in the private sector, are presumed to exhibit them, and other agents, say public sector operators, are presumed not to exhibit them, data should be available to reflect this distinction. Yet it is extremely unlikely that data sufficiently precise and reliable to satisfy all critics, will be available. There may be enough indications for us to make up our minds, and enough of these may be communicable to enable others to come to similar conclusions, but there will never be enough to confirm or reject the assumptions, and consequently to confirm or reject the hypotheses derived from that part of the body of theory that is supported by the assumptions.
Summarizing, our data will in all likelihood not conform closely to the variables in the theory, and if perchance they did, they will not be of sufficient number and precision to be wholly persuasive. Moreover, considering the fourth stage of investigation, the stern testing of hypotheses against evidence, this must be imperfectly performed. With few countries, few institutions, and a short time span, we are not able to provide enough observations to warrant any econometric analysis. Our statistical measures will be simple, as befits the data, and our conclusions tentative, as necessitated by the wide departure from the ideal.
As mentioned above, the next chapter will be devoted to theory, which provides the guide to our inquiry. Several theories will be considered, one of which generates hypotheses relevant to the issues we are studying. This most useful theory, presented as a model, will be covered in as much detail as necessary to instruct the reader; the others in much less detail.
In Part II, consisting of four chapters, one each on Ghana, Kenya, Tanzania and Uganda, we will present our data. The same format will be followed for each country: beginning with a brief recording of recent economic events and statistical aggregates; continuing with a report on individual institutions in the fields of agriculture, industry, appropriate technology and education; ascending to a compilation of resources devoted to advancing science and technology; and finally drawing a few implications for the country's future progress.
Part III contains the analysis. The data from the four countries will be assembled, together with other relevant information. To the extent possible the effects of Structural Adjustment Programmes on the pursuit of science and technology will be determined, first at the level of the entire economy (Chapter 7), next at the level of the sector of economic activity - agriculture and industry, exports and imports and non-traded goods, public and private activities (Chapter 8), and finally at the level of the individual institutions where scientific and technological tasks are performed (Chapter 9).
Part IV contains the conclusions, divided into two chapters, the first (Chapter 10) devoted to summarizing the results of the analysis in the preceding three chapters; and the second deriving implications for the future, both under the current regime of Structural Adjustment and under alternative, possibly more attractive regimes.
We now move on to the exposition of the theory, which not only suggests questions but also provides a framework which can support different programmes of Structural Adjustment, for there is little to be gained by trying to evaluate current programmes in the absence of alternatives.
2 Choice and application of economic theory
Introduction
Consequences, in theory, of advancing science
and technology
The timing, in theory, of advancing science
and technology
Theories of structural adjustment
Guidance
for our study
Economic theory can be put to many uses. The conventional use is the clarification of thought: specifically the explicit recognition of certain phenomena, their precise definition, the determination of their relations one to another, the elimination of logical inconsistencies, and the deduction of principles, often too subtle to be grasped intuitively. These principles, or hypotheses, can then be confronted with evidence and, if they are in conformity with the evidence, can be given some credence.
There are other, less methodologically pure, uses for economic theory. Theory can be used to buttress policy which was formulated previously in time or chosen for different, non-economic reasons. In economics, theory describes an Ideal, an environment organized in a simple fashion, containing elements identical in nature and following identical rules of behaviour. If the focus of the theorist is on the individual economic agents, and if the theory is powerful, it will generate interesting hypotheses about their performance and, perhaps, on the performance of all other agents of their type. If the focus is on the entire economy, the theory will generate interesting hypotheses on the performance of all the agents together. Whether the theoretical environment is small, containing just one agent, or vast, containing an infinite number, it describes completely and perfectly an economy: the economy is Ideal in the sense that its performance is predictable under the conditions specified. If the policy already chosen properly meets the conditions, then the performance of the system under the chosen policy can be predicted. If, in addition, the performance is an improvement upon the performance of the system in the absence of the policy, support for the policy is gained. The theory tends to justify the policy.
Theory is also used to contradict other theories. If an hypothesis, or set of hypotheses, is unattractive, an alternative theory may be formulated which yields, as its logical deductions, more attractive hypotheses. If one does not like one Ideal, one can create a better Ideal. Ideals are, after all, only imaginary representations of reality, and one's own imagining may produce a representation superior to that of one's predecessors. It has been said that much of the struggle of life is to persuade others of the correctness of one's own view of reality; the creation of theory is part of the process of persuasion.
Attractive or not, a theory can be made obsolete by a change in the environment that it purports to describe. If the changes are very slow to occur, or if the theory is particularly powerful and in conformity with contemporary ideological views, the theory may have a lease on life; in such cases the existing theory may coexist with a more recent one devised to take account of the changed circumstances. When the world contains actors operating under the old set of conditions and actors operating under the new, both theories can prosper. Methodologically, one should formulate a meta-theory, comprised of both sorts of actors, but this means of resolving dispute is usually unavailable: the meta-theory is either too complex to solve or can only be solved for specific, not general, conditions. One only has to imagine the difficulties of formulating a micro-economic theory governing an industry comprised of firms run by entrepreneurs with different motivations, or a macroeconomic theory describing an economy controlled by a government with inconsistent policies, to understand how formidable a task it is to develop meta-theories.
Nonetheless, even though it cannot in principle be said to be superior, the new theory formulated to take account of changed circumstances can be useful in suggesting directions in which research should move. This is theory in the role of guide to inquiry. ('Guide' is a modest word; often theory is used to justify undertaking inquiry.) The deduction from the newly-established theory is that such-and-such a phenomenon is important in the governance of a system; consequently that phenomenon should be studied. Its investigation is made respectable, is even made imperative, by the strategic position that the phenomenon holds in the theory.
Theories as examples of scientific method, as buttresses for policy, as contradictions to other theories, as guides to inquiry: these all represent practical uses. One final use must be mentioned, namely theories as aesthetic objects. Theories, like paintings or music, can be beautiful. 'Elegant' is the word used to describe a theory that is spare, balanced, well-defined and pregnant with hypotheses; and the formulation of such a theory can be a joy. It can give pleasure to the observer too, and may be assigned an importance because of the pleasure it confers. Like a sculpture carved by a member of a tribe that has disappeared from the face of the earth, a long-familiar theory may have an appeal that causes it to be cherished; in a museum of theories it may bear a proud place.
Having argued that theory can serve several functions, we would be remiss if we claimed that the only purpose to which we will put theory is the first, i.e. as an application of scientific method. We shall also display theory as the representation of an Ideal, as a justification for policy, and as a guide to our own inquiry, although not in that order. The first use, rather, will be as a justification for our focusing on the pursuit of science and technology.
Consequences, in theory, of advancing science and technology
The first use to which we shall put theory, as well as some statistical evidence, is in addressing two issues faced by all developing countries: is it worth while allocating to the pursuit of science and technology some of those so very scarce resources such as scientists, engineers and managers, capital, and time? The second issue follows upon the answer to the first: if it is worth while allocating scarce resources to the pursuit of science and technology, at what stages in a country's development should allocations be made? It is these issues that we shall use theory to explore.
Since the 1950s, much theory has been devoted to trying to capture, in the abstract, the effects that advancing science and technology may have upon a country's economy. The attempts were inspired by Solow's study (Solow, 1957), which indicated that only a minor portion of the economic growth of the USA could be explained by increases in the amounts of labour and capital applied to the production of the country's goods and services. The major portion was explained by a factor exogenous to the system described in the model, labelled by Solow 'technical change'. Technical change the factor surely includes, but not exclusively, for what the factor represents is everything other than labour and capital; it explains what labour and capital do not explain.
As might be expected, such a provocative statement as Solow's stimulated inquiries into just what 'technical change' signified. The response of Denison was, from our point of view, the most significant (e.g. Denison, 1967). What Denison did was to decompose the residual, left after the contribution of the growth of labour and capital had been subtracted, into several distinct elements plus, inescapably, a residual of the original residual.
Since Denison's decompositions covered the growth of the US and Western European economies they are of less interest to us than a similar decomposition of the elements explaining the growth of the economy of the Republic of Korea, carried out by Kim and Park (quoted in Westphal, 1986). In this rapidly developing country, between the years 1963 and 1982, increases in the supplies of labour and capital accounted for 55 per cent of the growth. The remainder, 45 per cent, is Solow's 'technical change'. How does this remainder break down? Kim and Park found that roughly 10 per cent could be assigned to increases in the amount of education gained by the workforce; 40 per cent to the attainment of economies of scale; nearly 20 per cent to an improved allocation of labour; and 30 per cent to unassigned causes, of which one involved was the greater utilization rate of capital equipment. The extent to which these different sorts of productivity increase can be attributed to the advance of science and technology in Korea is, of course, subject to argument; but, given the elastic definition of science and technology that we adopted in Chapter 1, we would be inclined to assign almost all.
Questionable an assignment as this may be, it does at least give some support to the proposition that advances in science and technology contribute to economic growth, and to our focus upon science and technology in countries eager to reproduce Korea's process of rapid economic growth from a low platform.