urbanization is a global process, pertaining to the spread of capitalism across the world, and Africa is no exception. Nevetheless, Sub-Saharan Africa holds a peculiar place in the urbanization movement among developing countries. From the 1950s to the 1970s, as opposed to what happened in other parts of the world, this region witnessed urbanization without industrialization. Then the urbanization movement slowed down in the 1980s and 1990s. Projections for the year 2000 show a 30% level of urbanization; that is, less than the average for all developing countries in 1980 and the whole world in 1950. The natural component of urban growth now prevails over the migratory and reclassification components in Africa. Two-thirds of urban growth was due to migration and reclassification in the 1960s, but only one-third in the 1990s.

But how can we analyse and understand such trends with the poor statistical systems that prevail in most African countries? What have been the changes in the statistical definition of urban areas, and do they fit the actual evolution of urban system in Africa? What definition(s) could help us to better represent the urban dynamics in Africa? These are some of the questions that this paper tries to answer, using various examples from Africa and comparing these with the situation in other less developed parts of the world.

Results: In all, 75% (N=207, F:M ratio: 3.6:1) of the eligible 276 subjects participated in the study. There was no significant difference between participants and non-participants regarding sex and living conditions. About half of the population were nursing home residents. In activities of daily living roughly one third was totally dependent, almost one fourth was relatively dependent and just about two fifth were relatively independent of help. Almost no centenarian could be described as being free from diseases. Cardiovascular disease was the most common disease affecting nearly 3/4 of the centenarians. Objective assessment of blood pressure and standard ECG revealed a substantial proportion of subjects with hypertension and myocardial ischemia, which were previously unrecognised. Disabling chronic conditions such as osteoarthritis, urinary incontinence and dementia were common, each affecting about half of the centenarians. However, at least one third of the centenarians were assessed as being non-demented. The presented disease prevalences can be regarded as minimum prevalences or "best case scenarios". Of major interest is the identification of a group, although small, of autonomous centenarians, who were non-demented, non-institutionalised and relatively independent regarding physical functions, and yet they had the same level of comorbidity as those who could not fulfil all these three criteria. Exemplified by the measurement of thyroid function, centenarians seem to possess some well-preserved physiological functions in spite of organ atrophy. Based on information from previous medical records centenarians were found to have suffered from several different diagnoses in the past, all potentially leading to a state of chronic condition and possible subsequent functional loss or even potentially mortal.

This paper argues the case for an urban system approach to the conceptualization, definition and analysis of the contemporary urbanization process. Cities develop, grow or decline through their relative position and functional roles within a larger system of cities. Those roles, indeed the spatial organization of urban systems themselves, are being redefined as the combined outcomes of the processes of economic competition, corporate restructuring, transnational migration and the globalization of culture, trade and investment.

using Canadian and uS examples, the paper illustrates the approach and documents its strengths and weaknesses, both as a conceptual framework and analytical tool and as a means of monitoring changes in urban places in an increasingly integrated world.

Two major challenges are currently facing contemporary metropolises. One is the challenge of the decentralization of the PUBLIC sector, where local intra-metropolitan governments are acquiring more responsibility for administering PUBLIC works and services, such as health, education and housing improvement, and are in some cases experimenting with direct citizen participation in the formation and implementation of PUBLIC policy. The second is a complex process of socio-spatial segregation that results in substantial differences between sub-areas of a metropolis in the income and educational levels of households, in the types of services demanded and in the quality of those services. The inter-connection of the two challenges is particularly clear in housing, where intra-metropolitan local governments face the issue of both upgrading recently developed irregular settlements, and of rehabilitating the now well consolidated but increasingly distressed self-help settlements that began 20-30 years ago.

To advance research in this area, it is crucial to identify data sets that are appropriate both for measuring socio-spatial segregation and for measuring the performance of local governments. At a start, it is important to identify data sets that can be geo-referenced to at least the level of local government units, and preferably to smaller units such as the equivalent of census tracts.

urbanization is a socio-demographic process in which societies are transformed from primarily rural to primarily urban ways of life. Hence, in post industrial societies urbanization cannot be understood without also examining the nature of rurality, and the extent and characteristics of rural community life. Perhaps it is axiomatic, but urbanization cannot proceed in post industrial societies unless rural communities persist and are at risk of 'becoming urban'. Some scholars (Brian Berry among others) have argued that metropolitan systems have become so extensive and so dominant in post industrial contexts that genuine rural life has ceased to exist. Scholars have also observed that globalization and modern telecommunications have further subordinated the rural periphery to economic, social and political forces originating in the urban core.

This paper examines the social and economic organization of rural life in post industrial society. using the united States as a case, we view rurality through a multi-dimensional framework to consider whether rural communities that are distinctively different from their urban counterparts persist in contemporary America. We present a comparative profile of demographic/ecological, and economic attributes of metropolitan and non-metropolitan areas during recent decades, and we propose how this comparison could be enhanced by the addition of information on institutional, and sociocultural dimensions.

We also examine the validity of the U.S. Office of Management and Budget's new 'core-based statistical area classification system' which is intended to distinguish between non-metropolitan counties that are genuinely rural, and those that are simply examples of the urbanized periphery. Our analysis will describe the demographic and economic composition and the institutional completeness of micropolitan areas in comparison with the territory outside core-based statistical areas, and the smallest metropolitan category. In essence, we will determine whether micropolitan areas are more similar to the smallest metropolitan category or to the territory outside of core-based areas.

Finally, we will comment on some of the reasons why rural communities and people continue to be highly valued in societies where rurality has been dramatically transformed from the modes of social and economic organization that existed when traditional pro-rural and/or anti-urban attitudes were originally formed.

This paper reviews the ideas and recommendations put forward by previous studies for representing settlement systems for the purposes of presenting population and related data. It looks at the recommendations issued by the united Nations Statistical Office, starting in 1948 and reviewed subsequently in the run up to each round of decennial censuses. It gives particular attention to the conclusions arising from the IUSSP's previous examination of this topic, undertaken by the Committee on urbanization and Population Redistribution in the 1970s under the chairmanship of Sidney Goldstein.

The underlying aim is to provide a marker against which the novelty of the observations and suggestions made in other papers can be judged. At the same time, where it is found that previous recommendations have not been widely acted on, it is important to discuss why this should have been the case. For instance, even as early as the uN's 1952 Demographic Yearbook, it was recognised that the settlement system should be considered as a continuum from small clusters to large agglomerations, yet the main element of the uN's reporting on population distribution trends continues to be the simple urban-rural dichotomy. Similarly, the Goldstein Committee's recommended developing consistent standards that would apply to groups of countries based on economic level and cultural background. It also advocated a multi-dimensional approach to disaggregating the settlement system, using the example of a four-way classification based on metropolitan/non-metropolitan as well as urban/rural distinctions. Why, therefore, does there appear to have been such limited progress towards producing international comparative statistics on a more refined basis? Whatever the reason, the mere fact of having to ask this question stresses the need for the practical viability and political acceptability of new ideas to be clearly thought through.

The complex processes which are re-shaping settlement patterns, particularly in advanced economies, create urban systems which cannot be captured by the simple framework of the long-established urban/rural categories. This complexity can be illustrated by the fact that place A can be more 'urban' than place B in some respects, but more 'rural' than it in others. The first task, then, is to identify key dimensions of modern settlement patterns, because it is these dimensions which have to be represented when moving on to devise statistical indicators of settlement patterns. Given this insight, existing settlement pattern indicators can be evaluated, leading on to the development of some more appropriate measures. The final step is to consider how several indicators can be combined in a multi-dimensional approach to representing settlement patterns.

Metropolitan phenomena in Brazil arose in the 1960s and 1970s when urbanization became more intense and led to a strong concentration of population and economic activity. In Brazil, where a lot of Metropolitan Areas are defined more in political/administrative than functional terms, it is possible observe a great diversity in terms of 'metropolization' processes depending on the region that one considers. While long-established MAs such as São Paulo and Rio de Janeiro saw an abrupt slowdown in demographics growth in the last two decades, others like Curitiba, Belo Horizonte (south and southeast) or Fortaleza and Natal (northeast) sustained accelerating growth. Despite this diversity, some characteristics are common to all of these areas: urban sprawl of the peripheral and poorest zones, intra-regional population mobility, building of middle and high class settlements in the outskirts etc. Rural population growth is another component of the demographic and spatial dynamics of some MAs.

using data from censuses and the Brazilian Household Survey (PNAD), this paper analyses regional differences in Brazilian metropolitan processes. It examines the potential and limitations of the data for addressing the aspects mentioned above. Special attention is given to rural population growth and to the possibilities of using approaches that go beyond the traditional urban-rural dichotomy.

This paper analyzes the development of the mega-city of Delhi from two interrelated perspectives: firstly, the evolving urban forms in the context of metropolization, focussing on the processes of 'peri-urbanization' and 'rurbanization', including expansion of suburbs, formation of new residential quarters in surrounding rural areas and the creation of satellite towns; and secondly, population redistribution within the metropolitan area.

This analysis is based on two main sources of data: decennial population censuses (latest one in 2001) and our own demographic surveys on population mobility (conducted in 1995 and 1996). The latter will allow us to illustrate the dynamics of settlement at the level of various selected zones in the Delhi Metropolitan Area, and according to the type of housing estate (planned sector, unauthorised colony, squatter settlement, urbanised village, etc.). In particular, we can evaluate better the respective role of migration from rural area or other towns in India, intra-metropolitan changes of residence, and absorption of villages into the urban area.

On the basis of this exercise, we draw some implications for categorising population for further demographic analysis (reflecting upon the inadequacy of a simple rural/urban dichotomy), defining relevant limits for measuring urban growth and delimiting zones for the purpose of town and country planning.

This paper describes and assesses the latest review of metropolitan area standards in the USA, the fifth since those issued before the 1950 decennial census. The united States Office of Management and Budget (OMB) issued its new 'Standards for Defining Metropolitan and Micropolitan Statistical Areas' in December 2000. This was the culmination of ten years' work, which - in addition to OMB's efforts - included research and testing by the U.S. Census Bureau, extensive review and guidance by a federal advisory committee that included representatives of six statistical agencies, and abundant comment from the PUBLIC, government agencies, and Congress.

All aspects of the previous (1990) standards were within the scope of the review, including whether metropolitan and related areas should be defined. Additional key issues in the review included the geographic units that would be used as the areas' 'building blocks,' the measures to be employed in grouping those units into statistical areas, and the frequency with which the resultant areas would be updated. This review produced new standards that are substantially simpler and therefore easier to use than the 1990 standards. At the same time, when applied with 2000 census data the new standards will provide areas accounting for an increased share of the nation's settlement pattern.

Two aspects of population dynamics are relevant to definitional/conceptional issues in the majority of More Developed Countries and are reducing the relevance of existing approaches to representing the settlement system for the purposes of analysis and policy making. One is that the metro/non-metro dichotomy is becoming less meaningful as an indicator of distinctions in the population. Secondly, within metro areas, the city/suburb dichotomy is also losing its discriminatory power. The classic role of (central) cities as destinations for rural-urban migrants and immigrants, who later move to the suburbs, is much less relevant today for a large number of metro areas, as entire metro areas are serving distinctly distinct roles as receivers of immigrants or domestic migrants. Similarly, non-metro areas are becoming less homogenous with respect to migration patterns and population characteristics. Gaining areas are serving as "exurbs" or "retirement and recreation magnets" and are sustaining in-migration of distinct populations (young couples, well off elderly). Losing areas are becoming older, and whiter as younger people move elsewhere. Finally, the old city-suburb distinctions in population characteristics are being eroded by the growing heterogeneity of the suburbs.

The paper examines these tendencies principally for the USA, looking at trends and cross-area differences in migration and population composition patterns. It also draws on uN data to look at trends in total population for the metro and non-metro areas of selected other countries. It discusses the main factors behind these changes, including the growth and geographical incidence of immigration, the eroding of old within-metro place boundaries due to increased within metro commuting and telecommuting, and the rise of a more "footloose" retirement population. It stresses that any new approach to settlement classification would do well to recognize the current and evolving distinctions that in many ways cut across the traditional ones.

This paper analyzes the evolution of the Mexican urban hierarchy since 1960. During this time it has evolved from a system of 119 cities with 14.4 million inhabitants (41.2% of the total population) to one of 360 urban localities with 64.7 million people in 2000 (66.3% of the national population) in 2000. The main features that it will examine are: the urban-rural population distribution by macro regions and States; the pre-eminent character of the urban hierarchy between 1960 and 1980; and the subsequent change into a polycentric configuration; the metro/non-metro evolution in the whole period; and, finally, the nature and characteristics of the emerging megalopolis of México City.

The paper includes an evaluation of the usefulness of the urban-rural distinction used in Mexico (localities with 15 thousand and more inhabitants are urban). It will also contain an assessment of the method of metropolitan delimitation used, as well as the pertinence of concepts such as sub-system of cities, polynuclear urban areas, citystates, nodal urban regions, and so on. use is made of locality-level data from the latest population census to consider the delimitation of metropolitan areas and the whole system of cities in Mexico.

This paper looks at how we define and understand rurality. Building on Halfacree's 1993 paper, it tracks the dematerialization of 'rurality' from seeing it largely as some kind of descriptive term to placing it firmly within the sphere of culture, as advocated by Marc Mormont and Paul Cloke, for example. A key issue to be addressed in settlement classification is the distinction between the material and the imaginative aspects of geographical concepts such as rural, urban, suburban, etc.

The paper also draws out the range of meanings attributed to 'rural' across Europe. In particular, it contrasts the importance of the term in Britain with its much lower significance as a device for our understanding of space in Germany. It suggests some of the reasons why 'rurality' remains an important concept for us to study when so much of the evidence suggests the ever-increasing erosion of urban-rural differences in advanced capitalist societies. It gives an indication of the deeper currents/ emotions/ intentions/ anxieties which can be read from the 'lure of the countryside', drawing on the results of research on different forms of counterurbanization.

It has long been recognised that where people live has an important influence on demographic patterns and processes. However, the spatial dimension remains weakly developed in most demographic analyses. Where the spatial dimension is incorporated it is usually in the form of a simple urban-rural dichotomy. It is argued in this paper that there are several dimensions of the settlement system which potentially can influence population processes and behaviour and that there is a pressing need to better articulate, conceptualise and operationalise these dimensions.

Going beyond the urban-rural dichotomy, the paper distinguishes three key dimensions of settlements - settlement size, population density and accessibility - and shows how each of these can influence demographic patterns and behaviours. It addresses some of the ways in which they can be defined and measured and the implications for data collection systems such as censuses. The paper also raises the issue of the spatial units of analysis employed in demographic study. These are usually restricted to official, often politically derived, regions which may have little meaning to people living in these areas. It is argued that more effort needs to be put into deriving systems of spatial units that are more meaningful in social and economic terms.

Finally, the paper argues that recent developments in geographical information systems (GIS) make it possible to both conceptualise and operationalise more complex dimensions of settlement systems than has been the case in the past. Moreover, the movement in national data collection systems to geocoding, extra emphasis on small area data collection and utilisation of non-traditional data sources greatly facilitate the recommendations in the paper.

In many Asian countries, both the levels and the trends of urbanization can easily be misinterpreted for two main reasons. The first is weaknesses in definition of urban areas. Definitions of 'urban' differ widely through Asia, but most use administratively defined urban areas above a certain population size, sometimes modified by additional criteria such as population density, proportion of male employment in non-agricultural activities, and possession of certain urban facilities. Factors working against modification of the criteria include inertia, and different financial arrangements for places defined as urban. The second cause of misinterpretation, which particularly affects large cities, is differential under-recording of the population in urban compared with rural areas.

From both planning and administrative points of view, one of the most pressing needs is to define appropriate mega-city regions focusing on the major cities. Boundaries of major administrative areas fail to capture the complex reality of peri-urban zones lying within and outside metropolitan boundaries. Approaches to providing more useful information on mega-city regions, and zones within these regions, are described and assessed.

This paper analyses and discusses the urbanization system in Latin America and Caribbean countries within the frame of population dynamics and development. It points to a number of deficiencies in our current knowledge of evolving settlement systems there. It goes on to suggest that this situation is highly related to the use of inadequate concepts and to limitations in the available population data collection systems.

The paper is organized in three parts. Firstly, it discusses the academic and policy importance to study the interrelations of human development and demographic dynamics in the Latin American context through the lens of the settlement system. The second part comprises a historical overview of population dynamics and urbanization trends in Latin America as a whole, with selected examples from national case studies. This includes two sections: i. contemporary dynamics of the settlement system, as far as can be seen from available statistics, and ii. a review and critique of current rural-urban classification schemes. Finally, the paper reviews suggestions for new types of settlement classification, including examples used in the region, and some recent applications of new technology applications.

Typologies of settlements are too often based on static reasoning. In order to improve the statistical description of urbanization, an evolutionary approach to settlement systems is suggested in this paper. An evolutionary theory takes into account the path dependency between urban and rural places. The concept of urban transition is included in the theory, which describes the transformation of a scattered, rather homogeneous settlement system made of small units into a much more concentrated and hierarchical urban system with agglomerations of considerably larger sizes.

Considering social interaction space instead of topographical space (through space-time convergence) is an essential explanatory element. The theory shows how variations in a few parameters like the timing and speed of the urbanization process can be connected with variations in the spatial and hierarchical organisation of urban systems which characterise distinctive types of settlement systems. Applications are made to Europe but also may help in predicting future evolution of settlement systems in developing countries.

urban places represent built environments that are physically distinguishable from the underlying/surrounding natural environment and are thus readily identified through the use of remotely sensed sources such as satellite images. This paper, however, goes beyond this simple identification of an urban environment to suggest that variability in the built environment is associated with variability in human behaviour in urban places. It is maintained that such variability in urban places can be captured through the classification of remotely sensed images and then analyzed within a geographic information system.

The paper begins with an overview of how satellite images can be used to distil information about the urban environment, followed by a discussion of the role that geographic information systems play in the analysis. This approach to understanding urban places is then illustrated using data from Egypt, in which variables derived from satellite images are combined with census data to improve our understanding of the spatial variability in human behaviour within the urban environment. Finally, suggestions are made of ways in which this type of analysis could be used to measure and understand phenomena such as urban sprawl and the multi-nucleation of metropolitan areas.

This paper examines changing urbanization patterns in China since the 1980s, especially the newly emergent in situ rural-urban transformation. It also explores related conceptual and definitional issues, such as the blurring of rural-urban distinctions in the urbanization process and the need for new approaches in defining China's settlement system and monitoring its evolution.

The paper starts by reviewing the overall picture of urbanization in China since the early 1980s, showing the increasing complexity of urbanization processes and inadequacy of the traditional settlement definitions for reflecting new realities. Then it examines in more detail the process of in situ urbanization, demonstrating various dimensions of the new rural-urban transformation pattern and the gap between the patterns of settlement evolution and the existing settlement definitions. Lastly, the paper discusses the implications of these trends for the modification of settlement definitions and for monitoring settlement changes.

This paper presents an overview of past and expected trends in urbanization by making use of the results of the 2001 Revision of World urbanization Prospects. It discusses trends and prospects regarding urban population size and growth, changes in the distribution of the urban population according to size of urban settlement, and the growth of cities with at least 750,000 inhabitants in 2000. It also analyses and discusses the availability of information on urban and city populations, the nature and limitations of the sources producing existing data, and the concepts underlying the statistics available.

Particular attention is given to an analysis of the set of definitions of urban and city populations provided by countries, their meaning and adequacy, and the criteria that are most often used to characterize urban areas. This provides a useful guide to actual practices regarding the identification of urban areas. It also aims to assist in the formulation of proposals regarding possible changes of definition that would make the resulting data more comparable or better able to reflect changes in the spatial distribution of the population.