Considérations relatives à certains aspects de la migration
Plusieurs organisations internationales s'intéressent de plus en plus à la dichotomie urbaine et rurale (la dimension «périurbaine»). Le Comité de rédaction de Réforme agraire, colonisation et coopératives agricoles examine l'analyse effectuée dans cet article, publié il y a 20 ans dans Migration et développement rural - sélection de thèmes aux fins d'enseignement et de recherche, (Etudes FAO: Développement économique et social no 3, Rome, 1979), particulièrement d'actualité étant donné ce problème «nouveau». L'étude de ce qui fait la différence entre deux points (distance) est présentée avec quelques propositions concrètes qui méritent d'être examinées par les décideurs en matière de développement rural, qui souhaitent véritablement éviter les effets de la migration. L'ouvrage porte sur le Programme de population et de développement rural réalisé par l'auteur pour le compte de la FAO avec l'Association des nations de l'Asie du Sud-Est (ANASE). L'ANASE comprenait alors cinq pays (l'Indonésie, la Malaisie, Les Philippines, Singapour et la Thaïlande), chacun étant chargé d'un domaine particulier. L'un d'eux a analysé les liens qui se nouent entre la migration et le développement rural. L'objectif des présentes considérations est de montrer comment ces liens peuvent être perçus sous des perspectives très différentes et avec autant de justifications, et que même les aspects qui paraissent le plus clairement définis (comme la distance entre deux points) sont complexes et dépendent des attitudes et des objectifs de ceux qui examinent la question.
Algunas consideraciones sobre aspectos específicos de la migración
Diversas organizaciones internacionales manifiestan un interés creciente por la dicotomía existente entre zonas urbanas y rurales (la dimensión «periurbana»). El Comité de Redacción de Reforma agraria, colonización y cooperativas estima que el análisis que se ofrece en este artículo, publicado originalmente 20 años atrás en Las migraciones y el desarrollo rural: temas concretos de enseñanza e investigación (Colección FAO: Desarrollo económico y social, No 3, Roma, 1979), es muy pertinente para la comprensión de un asunto «emergente» y nuevo. Se presenta un debate acerca de las causas que explican las diferencias entre dos puntos -distancia- y se formulan propuestas específicas que merecen la atención de los responsables de la adopción de decisiones en materia de desarrollo rural interesados en evitar los efectos negativos de la migración. Este trabajo se llevó a cabo en relación con el Programa de población y desarrollo rural, organizado por el autor para la FAO con la Asociación de Naciones del Asia Sudoriental (ASEAN). La ASEAN estaba formada entonces por cinco países (Filipinas, Indonesia, Malasia, Singapur y Tailandia), cada uno de los cuales estaba al frente de un sector. Uno de éstos se concentró en las relaciones entre la migración y el desarrollo rural. El objetivo de estas reflexiones es señalar que las relaciones entre la migración y el desarrollo rural se pueden considerar desde perspectivas muy distintas con igual justificación, y que incluso los aspectos que parecen más definidos (como la distancia entre dos puntos) son complejos y dependen de los puntos de vista y objetivos de quienes examinan la cuestión.
Chief, Population Programme Service, Women and Population Division, FAO
Several international organizations are increasingly interested in the urban-rural dichotomy (the "peri-urban" dimension). The Editorial Board of Land Reform, Land Settlement and Cooperatives considers the analysis carried out in this article, originally published 20 years ago in Migration and rural development - selected topics for teaching and research (FAO Economic and Social Development Paper No. 3, Rome, 1978) extremely relevant to this "new" emerging issue. The discussion about what makes the difference between two points (distance) is presented with some concrete proposals deserving consideration by those decision-makers in rural development genuinely interested in avoiding negative migration effects. The work was carried out in relation to the Population and Rural Development Programme organized by the author for FAO with the Association of Southeast-Asian Nations (ASEAN). ASEAN then comprised five countries (Indonesia, Malaysia, the Philippines, Singapore and Thailand), and each country took the lead in one area. One of these focused on the interrelations between migration and rural development. The objective of these considerations is to point out that the interrelations between migration and rural development can be seen from very different perspectives with equal justification and that even the aspects that seem the most clear-cut (such as the distance between two points) are complex and depend on the attitudes and objectives of those examining the question.
The objective of these considerations is to point out that the interrelations between migration and rural development can be seen from very different perspectives with equal justification and that even the aspects that seem the most clear-cut (such as the distance between two points) are complex and depend on the objective one may have in mind when examining the question and one's own attitudes.
The considerations illustrate the fact that in the field of migration interrelations with rural development, there are no simple deterministic mechanisms and therefore caution and flexibility in the use of knowledge are as necessary as an understanding of people's resistance to new ideas and of rural agents' objectives. One should also be wary of generalizations from previous successes or failures.
Facts in isolation have little meaning; they must be given a meaning by observation and analysis and this influences considerably the picture which emerges. Their significance depends both on the interpretation of the observer, as will be seen in the remarks on urban biases, and to some extent on the level at which one wants to examine them, as is illustrated by the role of cash. Finally, the translation of this knowledge into action tends to be influenced by the norms with which the entire question of the mobility of people is judged.
This section sets out to show that the question of migration is invariably perceived from an urban viewpoint and that this urban bias has far-reaching consequences. It can influence the structure of research or other activities and also the interpretation of the findings of studies2. In the latter case, this may not only result in the dissemination of incorrect information to rural development personnel, but can also lead to their making wrong decisions in the field - decisions which would, however, be correct within the framework of an urban-biased view of rural development.
In many ways it is an extremely ancient phenomenon. Throughout history the development of cities and towns has given rise to a sense of superiority on the part of urban inhabitants - an assertion that townspeople are somehow more worldly and sophisticated and that rural people have a lower status. This has become so deeply rooted in people's thinking that it is expressed in everyday language such as, for example, the use of the word "peasant" in a pejorative sense.
In the last 200 years the bias has been reinforced by the advent of industrialization and the fact that, as a result of the experience of the Western countries, industrialization has been regarded as the key to development whether this is taken to mean economic growth or development in the widest sense. In any event it meant the growing importance of urban areas with their industries and a strong orientation towards the Western type of urban world with all the implications of this.
Thus, with development being generally industrial and urban-based, the rural areas are perceived by planners and scholars in a supportive role and in many cases receive only residual resources for development. Even many measures which appear to be rural-development oriented are implemented because they are also of direct urban concern. For example, the building of dams provides water for irrigation but also provides additional power for city lighting and urban-based industries. This is also why in many cases rural development, which is concerned with overall living standards of rural populations, is confused with agricultural production, which is oriented primarily to national development (e.g. the growing of cash crops for export), although it is hoped that at the same time increased agricultural production will foster rural development. However, the respective roles and objectives of rural development and agricultural production need to be correctly perceived. Over the last few years the area of population and rural development (i.e. their interrelations) has indeed begun to be perceived and charted.
Thus, the forms of urban bias in national development most directly relevant to migration appear both in the spatial distribution of developmental resources and efforts and in the models of lifestyle and values upheld.
The foregoing remarks may lead us to expect that studies in this area will tend to focus on questions which are directly or indirectly relevant to cities.
The focus of studies. Although in developing countries the rural population generally constitutes more than 50 percent of the population, priority is generally given to rural-urban migration studies. Any rapid look through a bibliography of migration studies confirms this. Such studies examine the flow of migrants, their characteristics and their problems of adaptation in the urban centres, but tend to ignore the repercussions on the areas from which they migrate which are of particular interest to rural development planners. The basic difficulty is that these studies have a fragmented approach to migration and do not see it as a complete system. This means that planners have considerable difficulty in interpreting and drawing conclusions from the findings for rural development purposes. For example, studies show that one reason for migrating to cities is to further one's education. How can such a result be utilized for rural development? Does it mean that better schools in rural areas would slow down emigration? This is probably not the answer. To be of use for rural development such studies need to examine the conditions which have led to this result, the conditions under which a person would have opted not to go to the city to further his education or, if he did go, the conditions under which he would return to a rural area once this further education is completed, etc.
This type of partial approach resulting from the urban bias has serious drawbacks when one wants to consider the interrelations between migration and the sending areas. For example, in the case of male migration to cities, the studies do not relate this phenomenon to the change of status of women in rural areas or its effects on the modernization process, questions which are of direct interest to rural development. If the studies carried out up to now had been on a broader basis and had asked this type of question, they would be far more useful.
However, it has been realized that the focus of studies was unbalanced and that there was indeed a form of urban bias. Greater importance has been attached more recently to the study of rural-rural and urban-rural (or "return")3 migration. But these two areas have been studied in the same partial approach and the defects of this have resulted similarly in inadequate information. What is now needed is comprehensive and penetrating research in the original area of rural-urban migration.
The implicit value judgement. Studies of migration often reflect a desire to protect the city from an "invasion". There is often a basic assumption which is a value judgement that rural-urban migration is essentially an evil which should be checked. Such studies are concerned with avoiding the overburdening of urban services, a drop in their quality and the necessity of spreading investment resources which could otherwise be used exclusively for the existing urban inhabitants. However these studies often show an ambiguity because it is also realized that cities do need some immigration for their development and therefore the findings of the studies tend to encourage measures which favour migrants being selected according to the quantitative and qualitative needs of the towns, regardless of the consequences in the rural areas. It is therefore not easy for rural development planners to benefit from the findings of these studies. Moreover, their application to the urban areas might even lead to a widening of the gap between urban and rural areas and could theoretically aggravate the problems, actually increasing rural-urban migration which they implicitly aim to reduce.
A distorted perception of families and individuals. Because of the previous tendencies, migration studies often highlight the economic and particularly the cash aspects. The concepts elaborated and the tools utilized to observe reality are thus influenced. The primordial importance of the economic sphere in cities tends to be transposed to migration studies. But, in order to achieve a better understanding of the phenomenon a more refined approach is necessary.
For example, it is well established that many migrants come to cities in order to find jobs and this fits in perfectly with the economic approach. However, one has seen that migrants are not homogeneous: those who are educated are mostly familiar with the urban world and are very different from those uneducated ones who come for survival reasons (for instance, if they have become landless) or to earn cash to spend in rural areas. In these two cases the significance of employment and a cash income is not the same. There is more to it than just a difference in level of income between the groups; it also involves the meaning of money, particularly in its cash form. Suppose for example that the wages of these migrants were paid only in kind. How would this affect migration and the selection of migrants? It seems that besides wanting employment or the "city lights", migrants appreciate particularly the options opened up by the availability of cash. Cash gives them a choice of uses for food, entertainment, clothes, etc. This type of freedom is generally not available in rural areas and is not shown in the usual economic approaches. The urban bias does not enable us to take directly into account the non-economic aspects of the mode of payment which are connected more with this kind of difference between the rural and urban way of life. The role of cash in migration has not yet been explored in a systematic way although it is important in the understanding of the phenomenon of migration, especially as it is related to the emergence of the notion of a person as an individual economic unit4 which is part of the process of change from a rural to an urban way of life.
These remarks may seem somewhat abstract, but they do have a sound basis if they are examined further. For example, although the disappearance of the extended family has often been predicted it still survives in many forms, some of them disguised, and surveys may describe as separate households what is in fact one functional unit. This is a serious difficulty when carrying out economic and social surveys.
In a certain sense rural families function like a multinational firm: although individuals (like subsidiary firms) might appear to be disconnected, they still operate as a family entity. Many situations which appear irrational are thus made comprehensible, such as maintaining a lower standard of living than would otherwise be necessary in order to send home remittances (like operating at a loss) or decisions to send children out to the city (like creating a subsidiary firm). In this situation the uneducated individual migrant to the city only appears to be an individual, whereas the more educated migrant who leaves his rural environment and cuts the links with his rural origins operates much more as a real individual unit. The usual survey methods are geared to describing separate statistical units and the more educated migrant is therefore relatively well described. However, these methods give a distorted picture of the uneducated migrant by classifying him as an individual unit whereas he is, in fact, a component of the traditional network of relationships (extended or joint family, etc.).
Implications for action. Another example of the difficulty of interpretation and utilization of the findings of studies can be shown in the case of education. Studies always point to the effects of education on rural-urban migration.5 This is true, but what is the role of education? It is in fact a training for the urban type of life and therefore migration is a manifestation of its success. It is then said that this could be corrected by a more rural/agricultural oriented type of education. Should rural development encourage such a change? From the production point of view, most probably, but it may not have much effect on migration because it is not so much the topics covered in the education which are important, but the fact that the "education" normally promoted is part of a different world. In such a context a "rural education" would have the same philosophy as an urban one. Development has been criticized for being the imposition of a technical, impersonal, economic world on an older one often called "traditional", "archaic", etc. This is perhaps true, but it can also be seen as an attempt to facilitate the integration of urban and rural areas - that is, to close the gap that separates them. Expressions such as "extension" or "rural outreach" can be interpreted in both ways and probably reflect both aspects.
In spite of the foregoing discussion in which the differences between the urban and rural worlds have been stressed, they should not be perceived as essentially different and separate. Population studies in general and rural-urban migration studies in particular have created (partly for the sake of convenience) the urban-rural dichotomy. This fits in nicely with the "push" and "pull" analysis and has many advantages except perhaps that it might be misleading for rural development because rural and urban areas are interrelated. Besides the normal descriptions of the difference between rural and urban areas, one of the drawbacks is the implication that rural areas are homogeneous. Differences in population density, its spatial distribution and the location of areas (for instance, in relation to their distance from towns) tend to be overlooked. There are differences between rural areas themselves which might be important for their implications for rural development programmes. For migration studies to be really useful for rural development planning, a positive effort to promote in-depth studies of rural areas is necessary.
It is also a fact that the definition of "urban" varies from country to country, but most definitions tend to put any area which is not "urban" into the rural category, i.e. it is a residual category. This kind of definition also influences the usefulness of standards of migration for rural development. However, if one takes into account the variety of definitions and the fact that rural areas are far from homogeneous, one realizes that there are all kinds of intermediate stages between the most "urban" area and the most "rural" one. Added to this is the fact that many urban areas can have what amounts to rural pockets of population. Finally, the situation is far from stable; in most countries both cities and rural areas are changing rapidly.
The conclusion is that to be really useful for rural development policy, a migration study should:
Therefore one of the key variables as far as migration is concerned in the rural development context is the accessibility of goods and services. For example, since it seems that in certain kinds of rural areas (probably those not too far away from urban areas), providing amenities such as rural hospitals might accelerate out-migration, one must ask oneself why this happens. It is possible that the rural person, once he is aware of this amenity of the urban world, is no longer satisfied with what he might consider to be the second-rate quality of the rural hospital compared to the urban one. The rural development problem should then be reformulated from "Can we build (and should we build) small hospitals in rural areas?" to "Should one ensure that rural people have easy access to urban hospitals?" This question is based on the premise that one cannot distribute all the equipment and specialists evenly over a whole country. Should one have health and educational screening systems - for instance, via rural medical centres - which are a function of the degree of accessibility of the hospitals and schools? This kind of approach and reformulation of problems might have implications for migration. These implications need to be identified and quantified.
By creating the urban-rural dichotomy which, as has been seen, is a useful statistical convenience (up to a certain point), social scientists tend to fit reality into this qualitative classification. Persons or objects belong or do not belong to the urban (or the rural) category; there is no intermediate state and as a result the notion of the distance between the categories tends to be ignored. Therefore a person belonging to a rural area contiguous to an urban district receives the same treatment in studies as one far from any urban centre. Studies might look into the different characteristics of rural people but not into how far away, or how isolated they may be. The next section looks into this question in some detail and points out some of the implications.
Social studies might be more relevant to rural development problems if they discarded when possible the urban-rural dichotomy and used the notion of an urban-rural continuum which might be closer to reality and therefore more significant. Such a continuum has to be defined in each country because it changes over time and is influenced by cultural and developmental aspects. Furthermore, the concentrations of inhabitants in space also change both as regards their density and their location. In a three-dimensional representation they would look like a countryside with scattered hills (see Figure 1). In developing countries with primate cities, the configuration would resemble a mountain with some foothills and plateaux.
 |
Figure 1
A three-dimensional representation of the urban-rural continuum
However, in the present case it is more useful to represent the various types of urban-rural continuum in a simpler, two-dimensional manner (Figure 2). In this representation a country with most of its population in rural areas would have a distribution as in A. In a country with a primate city and no other important urban areas the distribution would be as in B and in a more balanced case it would be as in C. Therefore change over time tends to make the curves shift from A to C.
 |
Figure 2
A two-dimensional representation of the urban-rural continuum
The final shape and the shapes the curves go through depend to a certain extent on both the type of development approaches the country opts for and the impact of these on the related patterns of population growth and migration. It can be seen from these curves not only that general migration rates change over time but that the differentials in the selection of the migrants also change. Certain types of curve favour the differentials working more for skilled labour or unskilled labour, etc. The next section examines some of these aspects through the particular concept of distance.
The "traditional" approach is extremely simple since it involves three factors only: place of origin, place of destination and the distance between the two (Figure 3). The distinction between these three factors appears to be clear but, as this section shows, the distinction may be somewhat deceptive once one begins to discuss what is meant by duration of stay, change of residence, exact location of residence, and how these can be measured, if indeed they are quantifiable. This discussion focuses on what appears to be the simplest factor of all: distance.
 |
Figure 3
The three
traditional factors of migration
Let us take two rural areas, one close to a city and one far away and isolated. The agricultural production of the area close to the city is influenced by the proximity of the city (e.g. production of fruit and vegetables for city consumption, cash crops, etc.) and a rural agent has to take this proximity into account in suggesting changes in agriculture.
Most probably one of the tasks of the agent might be to encourage agricultural change in order to better meet the needs of the city. In a similar way the proximity of the city influences the rural population: some people might commute, others will opt for a try at city life. The mobility (i.e. the propensity to migrate) of people living near cities might be quite high, although it most certainly will be directed towards the city rather than to a remote rural area. The rural agent has to take such factors into account: it means, for instance, that in the case of families with older children, these will be working or studying in town and unavailable for work on the farm, etc.
In an isolated rural area, on the other hand, the agricultural patterns might be quite different, with little capital available for modernization and a concentration on subsistence farming. As regards the migration aspect, the situation might be the reverse of that of an area near a city: there will be little movement between the rural area and the town, i.e. there is a low propensity to migrate.
It is obvious from these two cases that measures taken to overcome the barriers created by distance (such as building a road to the isolated area in the second case) will influence both agriculture and migration, and that rural agents must take these into account. But is distance such an obvious obstacle which can simply be measured in kilometres, even automobile kilometres, as in developed countries?
Almost all authors agree that distance is a factor in migration.7 It is also agreed that it functions as an obstacle to migration, but, this said, the real problems start: What is really meant by distance? How does one measure it or what are the aspects of it which can be measured and how should we deal with those that cannot?
The approaches that are described may appear at first glance to be either overlapping or mutually exclusive. This is due to the fact that distance, which appears to be a simple "pure" concept, is in fact a complex one because it is always related to other factors. The confusion derives from the fact that researchers focus on different facets of the same reality. In the four approaches discussed, two cases attempt to partly synthesize the reality of distance, one by highlighting the physical aspect, the other by singling out the social aspect (the physical aspect is absorbed in the treatment of the social factors). In the third approach distance and other factors are broken down into a multitude of items and in the fourth approach a general synthesis is attempted by some authors. Our purpose is not to criticize the approaches in a general manner or to point out the strengths and weaknesses of the models; it is limited to analysing the way in which each one deals with distance.
Whether these approaches can be reconciled or not is discussed in a later section. For the purpose of this discussion we shall borrow heavily from several authors: Lee (1969), Courgeau (1970), Klaassen and Drewe (1973) and Amselle (1976). We shall distinguish the four approaches: physical, sociological, itemized and global-comprehensive.
The physical approach to distance. This is the most obvious one and is based on E.G. Ravenstein's benchmark work in 1885 on The laws of migration (quoted in Lee, 1969). He first pointed out the fundamental role of distance:
"The great body of our migrants only proceed a short distance ... [and] ... migrants enumerated in a certain centre of absorption will ... grow less as distance from the centre increases ... The inhabitants of the country immediately surrounding a town of rapid growth flock into it; the gaps thus left in the rural population are filled up by migrants from more remote districts, until the attractive force of one of our rapidly growing cities makes its influence felt, step by step, to the most remote corner of the kingdom."
As can be seen by the third part of the quotation, there is a strong analogy - an adaptation even, of Newton's law of gravity.8 This is confirmed (see Courgeau, 1970) by the formulation of the first models for migration where the exchange of people between two communities i and j is proportional to PiPj/r, where r is the distance between them. This basic formula has been perfected by modifying r with various coefficients and raising it to various powers.
The discussions around Pareto's model have generally focused on the significance of Pi and Pj and not on rijn.9 The distance and n have been determined empirically. For example, Hägerstrand finds that the value of n varies between 0.4 and 3.3. in Sweden. What is of interest here is that the lower values of n correspond to the migration of urban people and the higher values to those of the rural population. What is the significance of this?
Hägerstrand further observed that over a long period (one century) there was a general decrease in the values of n. This implies that the gap between rural and urban migration is closing and that the role of distance is diminishing.
Sociological approaches to distance. In such approaches physical distances are not formulated directly and other forms of "distance" appear. The emphasis is placed on "determining which variable, linked with distance would explain the distribution law of migrations" (Courgeau, 1970).
Let us see how "distance" might be represented in Hägerstrand's model:
Yij = kIDx/Pj
in which:
In fact I measures this by the previous migrations from i to j because it is measured as "a proportion of the total number of migrants from zone i having entered during a period of 15 years prior to t and Dx is estimated by the total number of migrants having entered j during the period".
From the rural development angle several remarks can be made:
These remarks show that both physical distance and sociological distance are important, that there are complex relations and trade-offs between them and that as a continuum, these factors can be manipulated so as to combine them in ways which could help the rural agent in implementing rural development policy.
An itemized approach to migration: distance as one item. Lee (1969) observed that "in the three-quarters of a century which have passed, Ravenstein has been much quoted and occasionally challenged. But while there have been literally thousands of migration studies during this period, few additional generalizations have been advanced".
In Lee's work, migration is defined broadly as a "permanent or semi-permanent change of residence". This leads logically to slight modification of Figure 2; the emphasis is placed on four sets of factors, i.e. those associated with the area of origin, the area of destination, the intervening obstacles and the personal factors (Figure 4). The factors associated with origin and destination are figured as "+" and "-" signs (these are not weighted), and the factors to which people are indifferent figure as 0.
 |
Figure 4
Origin and destination factors and intervening obstacles in migration
Lee also points out that this approach is an approximate one as the number of factors and the role of the personal ones, etc. make it impossible to be deterministic or hope for an exact representation. Because the pluses at the area of origin not only have to be more than the minuses at the area of destination, but also have to overcome both the intervening obstacles and the force of inertia which inhibits moving, this approach cannot be based on balancing out exactly the various pluses and minuses since a greater push factor is needed to overcome the force of inertia. However, it has the advantage of simplicity and flexibility. For example, in rural development the cost of agricultural machinery might be an intervening obstacle (-) which discourages migration at the place of origin if there is no hope of accumulating the necessary capital; at the place of destination the difficulty of obtaining a work permit can be a minus factor, and so on. One can break down the field of study into as many items as one feels necessary.
What happens to the notion of distance in such an approach? It should first be observed that distance is implicitly defined in a purely physical manner in Lee's model; for example, the psychological aspects of distance would be part of the personal factor, and the cost of transport would be another intervening obstacle, etc. This is already an important difference from the two previous models given as examples in which distance is not such a separate and isolated item, either from other intervening obstacles or from factors in the places of origin or destination. The major difference, however, is the fact that in view of the number of items, distance, although always present as an intervening factor, is by no means the most important one; for instance, laws restricting migration might be much more important.
Through his model, Lee establishes a set of hypotheses of which three are selected because of their particular relevance to rural development and its migration aspects:
It should be noted that these hypotheses lead to the same conclusion as the previous models, that is increased mobility of population. This is due to the fact that on the one hand development reduces the intervening obstacles - for instance, due to improved transport and communications; but on the other hand the differences between areas increase - for example, as regards the level of amenities and services available. The first part of this statement indicates that migration in developing countries can be expected to increase and that the role of distance should become less important so that it becomes a factor which rural agents could then ignore in most cases. To be effective in dealing with migration, they will have to identify in each case the other plus and minus factors and study the impact of their activities on them.
As can be seen, if one adopts the concepts in the first two models, distance is a factor which development agents should seriously consider, whereas in Lee's model it becomes so integrated in other factors (reality is so atomized) that its role tends to be insignificant. This is coherent given that the more factors there are, the less weight each will generally tend to have.
Now we shall look at a last approach which, instead of being an itemized one like Lee's, is a comprehensive one, and see how the concept of distance is dealt with here.
A global - comprehensive approach. This approach tries to place migration in a global and unifying perspective, both taken in the literal sense (Amselle, 1976). The level at which migration is analysed is therefore entirely different from the examples given up to now. In such an approach distance appears to a great extent to lose its meaning.
In order to understand this approach and its implications, it is useful to first summarize some of the criticisms of the classic analysis of migration made by the proponents of the global-comprehensive approach. This will then be briefly described and, in particular, its meaning in relation to the concept of distance will be explained.
Critique of classic migration studies. The criticism is twofold. First, it adopts a simple binary typology which concentrates on internal and external determinations in a narrow framework (see previous figures), whereas it is necessary to take into account the broad content which gives the coherence and unity to the multitudes of forms of migration. For example, how can one explain in certain studies the lack of a relationship between migration, density of population and size of farms? In the studies given as examples the authors explain this by the fact that those migrations were determined by the need to secure a cash income - a need which is related to the new forms of market economy which prevail in certain societies and which are independent of the local determinisms of factors of attraction or repulsion (i.e. Lee's pluses and minuses).
Second, studies explaining migration as a consequence of the greater mobility of certain societies due to internal factors linked to cultural patterns overlook the history of these societies and the change in their mobility.
Therefore, the critics of the classic analysis maintain, the narrow approach leads to an infinite fragmentation of the types of migration and this masks the real unity of the migration process which is comprehensible only in a global framework.
Although they do not refer directly to the problem of distance in their criticism, one can conclude that this is because it is irrelevant to some extent since it will either appear or not appear as a factor according to the type of binary typology adopted.
The concept of distance in a global-unifying approach. This approach considers for example that the world economy is moving towards a greater capitalist integration which favours the mobility of the work force.11 An example of such an approach can been seen in S. Amin's typology of rural zones in West Africa12 (see Amselle, 1976):
In this approach migration is not what it is usually taken to be, i.e. a movement over space, but is a change in social condition or status. Migration is, then, a form of socio-economic change accompanied by a movement in space. It serves to perpetuate and transform a society in the process of integration into the world economy and is thus a phenomenon which cannot be controlled at a regional or district level through rural development actions. Moreover, to some extent, it also defies control at country level.
It can be seen that in such a system, distance does not have much significance. It can even disappear as a concept in cases where the migrants maintain sufficiently good communications (remittances, flow of information, etc.) with their place of origin to attenuate somewhat their physical displacement. In rural development this would for example mean that the decision as to whether a road is built has an impact on migration only if the zone which the road would open up needs to be integrated, but not otherwise. If one wants to encourage migration, assistance to cover the cost of transport, etc. would have no significant effect under the assumptions of this approach.
It can be seen from the foregoing that the concept of distance, which appeared initially as a straightforward concept, is in fact rather more complex because of its relationship to other factors. Therefore, according, to the aspects one highlights, the logical implications for decision-making in rural development in order to achieve or avoid migration effects will be different. The whole area of migration and rural development is problematic and therefore merits much greater attention and research precisely because of the complexity of the relationships involved.
Therefore two simple approaches to distance are suggested here which will facilitate understanding. These are related but not entirely compatible and they are based on the fact that the findings of other research work are fairly similar, so that many of the apparent contradictions between them can be resolved when seen in another perspective. The approaches presented here have obvious limitations, insofar as only certain aspects can be quantified. However, they should be considered as only a preliminary attempt to cover new ground; there are still many difficulties to be overcome.
Is distance a continuum? Seen from the point of view of a continuum, distance is somewhat like a chemical mixture. It is not an element, as in each case the ingredients vary in their respective quantities but do not exist in isolation. The continuum goes from the most physical aspect of distance (e.g. walking hours) to the most abstract13 components of distance (e.g. differences in life styles or values) (see Figure 5).
 |
Figure 5
Representation of distance as a continuum
Let us now imagine the case of a city and two rural villages at different distances from the city. What are the respective distances of these villages from the city (see Figure 6)?
 |
Figure 6
Variations in the components of distance within zones and between zones
It can be seen that many combinations within a certain zone are possible. Because of this, measurement can only be approximate and rural agents will more or less know intuitively what that distance is. Figure 3 can also be used to represent the distances between rural areas themselves as well as differences in, for example, education between individuals which differentiate the psychological distance between each of them and the city.
Two kinds of rural development projects (electrification and road building) are used to illustrate the different elements of physical or abstract distance. Rural development projects such as these change certain elements in the concept of distance, either primarily from the physical or primarily from the abstract point of view. A road reduces the physical distance whereas electricity, by facilitating the advent of television, for example, can reduce the abstract distance. One thus moves from the concept of distance to that of accessibility. Consequently, these projects change the composition of the elements making up the total distance and they therefore have differential effects on the decision to migrate according to the characteristics and predisposition of the population in question.
It can also be seen that the perception of the distance between his place of origin and the city changes when a migrant goes to a city. He may well visualize a new abstract distance to his place of origin, as he gets adjusted to the city. This might have repercussions on his decision to remain in a city or on his difficulty in adjusting to his place of origin on return. By extending this idea, it is apparent that the introduction of an urban-type education in a rural area has the effect on youth of reducing the abstract distance to the city but of creating a psychological gap, or distance, between them and their place of origin or area and which affects their integration into this area and their propensity to migrate. This leads us on to a different approach to distance.
Is distance composed of divergent factors at several levels? If one reconsiders together the various approaches to distance reviewed, one can make several observations:
The continuum approach is therefore insufficient. For example, one can immediately raise the objection that if, for example, one wants to attach more importance to the abstract distance, one must at the same time move further away from the physical distance pole. At least a two-dimensional approach to distance is therefore necessary and the possibility of a divergence between the social and physical distance must be taken into account (Figure 7).
If one uses the notion of abstract distance one can represent distance as in Figure 7 in which 1 represents the presence of many obstacles, both physical and abstract - a total barrier; and 0 the absence of any barrier factor. As a result there would be three types of combinations of factors for each group or individual: combination C for which the values of distance are too high for migration, i.e. there is a combination of many obstacles affecting the potential migrant's perception of the abstract distance, plus many physical obstacles; combination B in which there is a choice; and combination A in which migration would be probable, as both the physical and abstract distance are minimal.
 |
Figure 7
A two-dimensional representation of distance
In such an approach rural development measures can be decided upon. Those which would encourage migration would mean a move from combination C to B - for example, taking measures to overcome the physical distance: building a bridge, reducing the hazards of the journey; and likewise, to reduce the abstract distance, information and motivational activities to make those in combination B decide to opt for migration (and vice versa if the objective is to reduce migration).
This representation leads to a more differentiated and subtle approach to development measures in each case. Not only does this minimize the danger of imposing a blanket solution but also it becomes possible to combine various approaches and to study their compatibility or to programme them - for example Phase 1: reduce the physical distance; Phase 2: motivate those in combination B, and so on.
Perhaps a word is necessary to justify the combination B zone of choice on whether a move is made or not. This is based on the assumption that migration "laws" are never fully deterministic. They only increase the probability of deciding to move and this probability is represented as an area of choice which is, on a general level, undetermined; whereas in zones A and C it would require a strong decision in view of the situation to either move (combination C) or not move (combination A) when the factors determining distance would point logically to the reverse decision. For example, there are always the people who would move, or refuse to move, whatever the circumstances.
If the representation in Figure 7 seems satisfactory for the first two points set out above, it does not answer the third, i.e. the problem of level. In order to do this a third dimension has to be introduced (Figure 8). For lack of a better expression we use here "change of civilization" to express the degrees in this third dimension.
 |
Figure 8
A three-dimensional representation of distance
In order to explain this simply, one can consider a few examples: a move from one rural area to another rural area generally does not imply a basic change in a migrant's mode of living. The farmer probably remains in the agricultural sector, although the crops or tools might change. The difficulty of learning how to grow a new crop is part of the abstract distance and the number of days' travelling is part of the physical one. This implies that often there is no movement to a rural area if the advantages in moving are considered too small in comparison to the distances. In most cases, however, the agricultural population knows the type of problem involved in overcoming such distances and their interpretation of rural development measures and their implications for such a move can be discussed and more easily understood by rural people. These matters relate, for example, to yields, amount of work, forms of ownership, problems of irrigation, etc. But a move from a rural area to a city generally involves more than a change of lifestyle - it is a change of civilization, of world even. It means moving from an agricultural world to an urban one with different norms and values. How does this new dimension operate in practice?
It has been observed that rural to rural migration tends to be undertaken by groups rather than by individuals, whereas in the case of rural-urban migration the reverse is true. This can perhaps be attributed to the fact that the individual who migrates to the city is very likely to be educated and it is the very nature of his education that has set him apart from his group and made him into an individual unit prepared to go the rest of the way to complete adaptation to the Western, urban lifestyle by migrating to the city. In contrast, people who have not received this reorientation through education stay within the group, which therefore remains traditional and migrates to the areas which are the least strange to it.
By migrating, the individual is in effect integrating himself into the world he really belongs to and leaving the one from which he has been alienated. Development in rural areas often constitutes an attempt to change the environment or at least to introduce into rural areas elements of the urban world; as a result, the people who accept these changes most easily may tend to go straight to the city rather than wait for further change to come about in the village. Conversely, there are practically no elements of the rural world in city life and this might explain why it is so difficult to encourage city dwellers to move to rural areas. This is all the more true since in developing countries the urban world is generally taken to be superior to the rural one.
However, it has also been observed that uneducated rural people move to cities and, even after a few years' stay, return to their place of origin little affected. In such cases, these migrants have probably lived in the cities with people of their area of origin and in fact have not really come out of their rural world at all. The third dimension and some of the abstract dimensions are thus reduced to a minimum. This is one of the reasons why many towns in developing countries do not have all the accepted urban characteristics, but are analysed as an agglomerate of villages.
It has also been observed that there is a stepwise migration: from rural areas to small towns, then to larger ones, then to primate cities. This stepwise migration is generally undertaken by uneducated people, whereas educated rural youth cover greater distances and tend to skip some of the steps. This could be explained by the fact that the uneducated rural migrant, by going to a small town tends to reduce the element of change of civilization incorporated in the distance. He settles down after each change before moving on to the next step.
The third dimension applied to the previous examples can also help us to understand why, when the physical distance is the obstacle, migration tends to be more permanent, whereas when it is the abstract distance, and especially the change of civilization, which is the obstacle, migration tends to be temporary. It will be remembered from the first model presented that Hägerstrand observed that the value of the power modifying the distance diminished over time and was smaller for the urban migrants than for the rural ones. One can thus interpret these values as representing the third dimension just introduced. For example, over a long period in Europe, the countryside has been conquered by the urban civilization model: this is translated by a reduction in the value of the power n already discussed. In a similar way this also explains why mobility increases with development.
Finally it is perhaps this third dimension which explains why rural-urban migration, even if urban populations seem to have swollen to enormous numbers, actually affects only a small proportion of the rural population: the distance to be covered in changing from a rural to an urban world is considerable and discourages most people. By the same token it shows that, if not properly managed, migration can get out of control if the distance is reduced, and this consequence is generally overlooked when rural development activities are being planned.
1 Originally published in Migration and rural development - selected topics for teaching and research, FAO Economic and Social Development Paper No. 3, Rome, 1978.
2For a general study of the urban bias (although from a different perspective) see Lipton, 1977.
3I.e. the return of comparatively recent rural-urban migrants to a rural area.
4As distinguished from the person in the rural world, who is defined more within a system of family, religious, ethnic and other group relations.
5For an approach aimed at eliminating some of these effects, see FAO, 1969.
6The aim of this section is also to encourage research. As the discussion on distance points out, there can probably be no universal model and many of the components of distance cannot be quantified directly. However research carried out in the countries can perhaps find ways of arriving at estimates of the components relevant in the local context which would be sufficient to satisfy local planning and implementation needs.
7However, manuals on demography such as the United Nations' (1971) Methods of measuring internal migration and Shryock and Siegel's (1971) Methods and materials of demography hardly pay any attention to distance.
8Any two bodies in the universe attract each other with a force that is directly proportional to their masses and inversely proportional to the square of their distance (d) apart: F = Gm1m2/d2.
9rij is the distance between Pi and Pj in each case.
10The weights would also need to be adjusted to the different types of migrant.
11Note that the practical conclusion, in a different formulation, i.e. an increase in migration, is similar to the previous models. Klaassen and Drewe (1973) also consider the relation between the integration of economies and the impact on mobility as a fact.
12Here also note that this to some extent recalls Ravenstein's laws.
13Abstract distance is defined as a combination of social, psychological, economic and other factors which influence the migrant's perception of distance.
Amselle, J.-L. 1976. Aspects et significations du phénomène migratoire en Afrique. Les migrations africaines, p. 9-39. Paris, France, François Maspero.
Courgeau, D. 1970. Les champs migratoires en France. Travaux et documents No. 58. Paris, France, Institut national d'études démographiques.
FAO. 1969. Planning for better family living. NU: Misc/69/28. Rome.
Klaassen, L.H. & Drewe, P. 1973. Migration policy in Europe - a comparative study. Westmead, UK, D.C. Heath Ltd.
Lee, E.S. 1969. A theory of migration. In J.A. Jackson, ed. Migration, p. 282-297. Sociological Studies No. 2. Cambridge, UK, Cambridge University Press.
Lipton, M. 1977. Why poor people stay poor. A study of urban bias in world development. London, UK, Temple Smith.
Shryock, H.S. & Siegel, J. 1971. Methods and materials of demography. Washington, DC, USA, Bureau of the Census. 2 vols.
United Nations. 1971. United Nations manuals on methods of estimating population, Manual VI, Methods of measuring internal migration. Population Studies No. 47. New York, NY, USA, UN Department of Economic and Social Affairs.
