Theories of Urban Dynamics

This paper reviews the existing analysis framework for territorial dynamics and urban growth and proposes a taxonomy of interpretive theories as well as a critical review. Specifically, the paper aims to provide four innovations to existing knowledge in this field as follows: firstly, a clear presentation of how the data of population growth of each habitat type have appeared and their academic interpretations; secondly, a reclassification of interpretative theories into three groups: the counterurban, the postfordist, and the cyclical theories; thirdly, with the ultimate goal to analyze the suitability of interpretations to the reality, a taxonomic proposal of habitat categories being made; fourthly, the final one refering to the balance of the theoretical to the empirical reality, in light of the data currently available, using the considered methodologies. That balance allows collecting positive elements of each theory and pointing to the possibility of developing a theory of synthesis.


Introduction
The study of what happens to human settlement patterns, from a demographic perspective, had a difficult acceptance in the academy.It has been tackled mainly from two disciplines: the study of internal migration and the local development.However, other areas have greatly contributed to the generation of an enormous literature on the subject but have not led to a general consensus theory [1].The decade of the seventies of the twentieth century recorded a remarkable increase in these studies due to the extending of some phenomena such as suburbanization, urban deconcentration, or counterurbanization.
The literature published in the last decades has been very contradictory, as well as the statistical data published at the time.Interpretations and reinterpretations of the demographic interterritorial dynamics changed repeatedly their sense several times, even for the same analysts.Currently, we are able to analyze from a more distant time perspective the processes that were going on.Even it is possible to check out which one of the predicted trends turned out to be more accurate.With this intention, the expository sequence of this text is structured in four steps: it begins by setting out the chronological presentation of the analysis, the data, and their interpretation; then, the categories of relevant habitats are proposed to analyze the territorial dynamics; thirdly theories that have emerged and still survive are classified to explain these events; finally, a balance of the right and wrong answers of each theory is performed (chronologically, each decade has shown demographic patterns opposite to the previous trends.This has led to conflicting interpretations.The differences between them are based on the answer given to three questions: what happened, why, and what will happen.Different types of theories appear as a response to these issues.The relevant habitats to understand the dynamics are also three: the city, the periphery, and the rural.Each of them can increase or decrease demographically, regardless of what happens with the other two).
The main hypothesis defended in the paper suggests that, in general, existing theories have insufficiently explained the population dynamics between different habitats.In fact, they all have right but also wrong answers.The possibility of the emergence of a theory of synthesis is a task not resolved in this text, but viable.

Changes in the Dynamics of Urban Growth
In recent decades, several changes have been observed in the dynamics of growth in cities that questioned the classical 2 International Journal of Population Research paradigm of urban concentration.According to this, the world is inexorably heading for a process of increasing urbanization, which implies population concentration in the densest cities, to the detriment, fundamentally, of rural spaces.This was an evolutionist interpretation, understood as a process of social change, tied to another broader process, related to modernization and industrialization.
There are no doubts about the historical and temporary relevancy of this urbanizing trend.In this way it would be necessary at least to characterize urban development tied to the initial phases of industrialization.On the other hand, from the end of the twentieth century until our days, data shows that this paradigm requires certain modifications.As Berry and Beale pointed out early on, large metropolises first showed symptoms of depletion in the USA in the 1970s.The same trend also became perceptible in the most developed countries in Europe soon after.This initiated a period of huge counterurban boom, when deconcentration was linked to both modifications in the productive structure (including transport) and in individual preferences, always under a suspicion of urban crisis and decline.Some analysts even understand the city as a relic of an obsolete industrial society, replaced by a new postindustrial society that was generating a different territorial structure.Storper and Manville synthesized this argumentative framework in the following words.
"The cities were the jetsam of another age, vertical settlements in a horizontal world, artifacts of a time before distance died.They were not where people wanted to live and were no longer where they had to work" [2, page 1248].
The change from the first phase, urban concentration, to the second one, consisting of deconcentration and turnaround, was seen in the academic world as a change of enormous scientific and social importance.Strauss [3] concluded that an authentic "clean break" within the past had taken place.Nevertheless, these first interpretations are slightly confusing, fundamentally because of the gaps in the delimitation of the exurban (extensively understood as rural, or simply as "not metropolitan"), including small, rural towns integrated in urban regions or even clearly peripheral spaces in cities.A debate then arose on the field where the calculations of balance and growth should take place, under the suspicion that certain "statistical illusions" had been generated, the product of erroneous definitions of "urban." To the detriment of the turnaround and the "clean break, " since the mid-1980s, when evidence for deconcentration was most consolidated, more recent data started to show signs of recovery and urban revitalization in general.The evolution recorded later by the main cities soon put an end to the counterurban optimistism.After this data from the 1970s, the following decade saw North American metropolitan areas gaining more population than nonmetropolitan areas.From 1980 to 1990 North American nonmetropolitan areas gained 2.7% as opposed to 11.8% in metropolitan areas (Bureau of Census).The reversion of the eighties also extends to many other western countries.The "turnaround seemed over, destined to have been a short-term aberration" [4, page 1355].The deconcentration trend of the 1970s was interpreted from certain positions as an exception to the more general urbanizing trend.Among its candidates the idea that cities had periods of "fat cows" and others of "lean cows" became generalized.At that time, history was recording a period of crisis from which the city rose again to impose itself as the dynamic habitat par excellence.
When everything seemed to be in order, however, from the more prourban theoretical perspectives in the 1990s, small towns recorded much more favourable balances than large cities, at least in the most industrialized societies.A nonmetropolitan revival in the USA was perceived halfway through the decade.In other areas, diversity became evident and countries that had previously recorded urban crisis changed and once again showed clear evidence of concentration; nevertheless, there were any others that showed symptoms of deconcentration.There are currently two different interpretations: on the one hand, the hope of the counterurban theorists who visualized a return to the 1970s, interrupted in the previous decade, but understanding this regression as the last "death throes" before the definitive and fatal crisis for cities, and on the other hand, the insistence on questioning the counterurban interpretation, insisting that what grows is mainly and in most cases cities' outskirts and suburbs.Furthermore, and to complicate the model, a process of regeneration for inner cities is stated, giving rise to what was named reurbanization from that time on.
In the late 1990s, the net migration flow between US nonmetropolitan and metropolitan counties, which had been favourable to the former in the first half of the decade, changed and became favourable to the main cities.This was soon detected by Beale [5] and Cromartie [6], while Johnson et al. [7] later ratified the trend.A new trend thus came to light in favour of the prourban thesis.
In the new millennium the proof of reurbanization became clearer, both in America and in Europe [8][9][10][11].The adaptation to the cyclical model seems quite acceptable [12], although, from certain positions, there is an insistence, now clearer, on the simultaneity of certain processes [11,13,14] and even on the tendency towards stabilization in all habitats [15].
Historical and empirical analysis has enabled us to state that the series of changes from the "clean break" went through the following sequence (Table 1): until the 1970s, urban concentration in cities, accompanied by processes of suburbanization.From the early 1970s, a phenomenon that could be called "nonmetropolitan turnaround" is observed, characterized by population increases in nonmetropolitan areas and even by migration in the same direction.This phase was followed by "turnaround reversal, " which was dominant in America in the 1980s [16,17] and where growth in nonmetropolitan areas slowed down considerably and the migratory balance changed in favour of metropolitan areas.But in the 1990s, the trend changed once again, recording a "rural rebound" [18], in which, in general, "the rates in nonmetropolitan areas were higher than those in metropolitan areas" [19, page 2].Finally, in the first decade of the twenty-first century, the tendency towards a "Rural Rebound Reversal" [1] seems to have changed again and is now more similar to the 1980s pattern.

Interpretation
3.1.Levels of Analysis.The explanation and formulation of interpretative theories for these modern events have been broad, diverse, and somewhat confusing, in tune with complex and "erratic" territorial dynamics [1].Some decades after the pioneering formulations of Beale and Berry, it is now possible to try and put some order in this enormous literary production.The task is neither original nor pioneering, since there are significant precedents [20][21][22][23], but it does involve a novel approach due to the conclusions reached in this study.
How have scientists and scholars interpreted these changes in demographic behaviour in cities?The answer to this question involves understanding the specific features of the "mode" of the scientific approach to the topic.It often has to be found in "tacit" references, where no theoretical interpretation is explicitly expressed, in demographic studies, which are highly empirical and local, while, on the contrary, when it is explicitly expressed, it is aimed at three different, although complementary, levels of attention: what grows, why it grows, and what will grow in the future (Table 2).The distinction between these three levels (which we will call descriptive and causal interpretations and prognosis) is relevant because each of them follows a different logic, although interrelated with the other two.In "tacit" cases the answers to these same questions also turn out to be latent.
Descriptive interpretations of what happens (what grows) can be divided into two groups of answers: on the one hand, continuists or evolutionists and on the other, rupturists or revolutionaries.Continuists understand that the only thing that takes place in deconcentration is a change on the territorial scale; the city maintains its living dynamism but includes territories (urban areas) that go beyond the perimeter of the "continuous-compact" to include both spaces in "low density cities" and housing developments in denser estates.Rupturists place the emphasis on the fact that something relevant has changed and that cities have lost their historical hegemonic roll in favour of more dynamic new locations; we are faced with a "new cycle, " which arises out of the end of a previous stage (Berry, Vining, Kontuly, and Ascher).
Causal interpretations, on the other hand, are explicit in mentioning the reasons that generate the processes observed in said situation.One of the most exhaustive descriptions of the explanatory causes of territorial dynamics was drawn up by Bierens and Kontuly [20], who divide the positions into five groups, each one catalogued by an epigraph: regional restructuring (in which the key factor is the relocation of companies and their successes or failures), the perspective of deconcentration (that lays the emphasis on individual residential preferences), periodical effects (explained by the existence of expansive cycles for cities and a temporary crisis provoked by their saturation), governance (noise, crime, cleaning problems, etc.) and, as a related variant of the latter, the cost of housing (prohibitive prices and/or deficits in the supply).
Prognosis, finally, requires taking a stand with regard to the cause, albeit tacitly, although far from being a game of divination it fundamentally attempts to detect the structural inertia in social change for territorial dynamics.These dynamics can potentially establish 8 hypothetical scenarios, although they have usually been summed up in 4: counterurbanization, reurbanization, those who adhere to post-fordist diversity, and those who think that all spaces have similar dynamics.The matrix shown in Table 3 explains each of these 8 hypothetical scenarios of growth dynamics (+) or crisis (−) for each habitat (actually repeated permutations with 2 elements (growing or decreasing, symbolized in Tables 1 and  3 by + and −), taken for each habitat (there are three: core, ring, and rural).If  is the number of elements and  are the cases we use, the calculation of the possibility is   = 2 3 = 8).We consider every hypothesis mathematically possible.It is important to be exhaustive in the identification of theoretical alternatives, so that we can methodologically justify that the contrast is carried out for all possibilities.
Four types of prognosis for the future stem from this matrix are as follows: (a) the counterurban hypothesis of city crisis with the growth of rural and/or peripheral spaces (see situations numbers 5, 6, and 7 in Table 3); (b) the prourban and reurbanization hypothesis, including the presumptions of urban cycles that lead thereto (situations 2 and 4 in Table 3); (c) a mixed hypothesis, in which some urban, some rural, and/or some peripheries grow (any situation in Table 3 is partially possible); and (d) a rebalance hypothesis, after a territorial transition (situations 1 or 3 and 8 with nuances).
The considered habitats (compact city, periphery, and rural) are understood here as conceptually open, allowing different and alternative definitions.It is important not to close the definition of each one of the habitats (something that would have to be realized by criteria of density and contiguity) in order to give content to a major diversity of interpretations on them.On the other hand, it is indispensable to understand that these are three types of relevant spaces to interpret the territorial dynamics [24].Between the classic studies that develop these concepts and variations among different countries, it is necessary to mention van den Berg et al. [25], Hall and Hay [26], Johnson et al. [7], and Turok and Mykhnenko [11].
The relation between taking a stand for a variable and with regard to the rest is diverse, but there is a certain predominant logic that enables us to synthesize the paradigms shown in Figure 1.The classification of interpretative paradigms is a result of combining all these interpretations.Among these positions, there are multiple possible combinations, but the logical sequences and contradictions limit the possibilities to three main groups or paradigms: counterurban residential preferences, post-fordist productive restructuring, and theories focused on cycles and transitions (Figure 1).

Individual Counterurban Preferences.
A consolidated and frequent academic tradition considers that a city crisis was recorded with a negative correlation between migratory balances and the urban dimension.Beyond this few generalizations can be attributed to this broad but also diverse group, whose members are labelled externally as ruralists, researchers of counterurbanization, or simply theoreticians of deconcentration.
In general, it can be affirmed that they maintain a critical perspective with traditional compact cities.In contrast to the classical 1931 figure of Louis Wirth, Ruralism as a way of Life emerged with intensity in the 1970s and the term "chaotic" was institutionalized, a term that Berry christened as counterurbanization.Studies such as those by Mitchell [23], who makes a review and classification of the interpretations that appeared in the last third of the twentieth century, together with Sant and Simons [27], Dahms and McComb [28], Ferrás Souto [29], and Halfacree [22], among others, enable us to frame their points in common and their differences.
It is possible to conclude from these studies that the description of the process should distinguish (a) those who understand it as flow (migrations) as opposed to those who speak about balance (structures); (b) those who do not wish to limit themselves to rural areas [30,31], as opposed to those who believe that this should be the centre of attention [3,22,[32][33][34].Some of them even think that counterurbanization recorded different phases that have modified the destinations of the flows [35].
On a causal level there are two types of arguments: individual preferences for smaller habitats and the processes of productive restructuring.The interpretation based on preferences believes that the urban crisis is generated by a change in the tastes of the population, who generally choose to reside outside the inner city, in smaller, "natural, " and calm habitats.The model of the large city, dense and compact, is shown pejoratively, as the opposite of the rural, which becomes an idealized space of greater well-being and quality of life.In this line of argument we can locate the early Berry, and especially Vining and Strauss, who formulate this more ruralizing counterurban theory, demographically and culturally.Residential preferences were studied early on in the processes of decentralization by Fuguitt and Zuiches [36] and others.The references then get multiplied and there are numerous case studies.The conclusions point in several directions.Fuguitt and Brown [37] underline that quality of life is adduced by those who prefer small settlements as opposed to economic motivations (and especially wages) alleged by those more inclined to cities and urban areas.Nevertheless, Allen et al. [38] find that the majority of residents' social-demographic and economic variables are not significant.
Complementarily, other authors such as Butzin, Frey, Fielding, Audirac, and White insist that the cause of counterurbanization and the population's rural preferences are the result of changes in the productive structure.Production can be relocated, ways of working that do not require a specific physical location come into play, and above all, the possibilities of communication among people distanced in space increase.All this has been called productive post-fordist in such a way that some of the counterurban theoreticians share causalities in a similar way to "post-fordism." At a predictive level, they defend that these low density spaces will at least maintain their strength, although not exclusively.In general they have been characterized as defendants of the future of rural spaces, either as renaissance [34] or as the end of the loss of their population [39,40].Nevertheless, this prognosis has lost prestige in recent years.The reason is that the data from the 1980s and from early time in the new millennium does not allow us to clearly affirm that such a process is taking place and they even doubt that it ever took place.This has led to pointing out prevalence in their presumptions in the territories that have developed as destination centres for the elderly, tourists, or others [22,41,42].

Post-Fordist Productive
Restructuring.The argument that territorial dynamics are the result of processes of structural change, that they refer to changes in the typology of employment, the location of work centres, the best communications, and an increase in the diversity of residential formats, is highly recurrent.The theoretical presumptions that group them together argue that the kind of habitat is not the most relevant variable, since there are cities that are immersed in a deep crisis, while others adapt to the demands of the new society; likewise, some rural spaces find their strategies for surviving while others are deserted.The explanation for why this happens can be generalized in the common theoretical paradigm of post-fordism, understood as flexibility, the adaptation of products to markets, and in general the absence of uniform and universal patterns in the framework of new structural and productive changes.
In practice this paradigm is linked to everything that is related to a change in the model, characterized by uniform and chain production, to another more flexible model, with devoted workers and products adapted to the consumer.In its territorial version it is understood as the change in the process of population concentration in the main cities, to another process involving the deconcentration of population and functional adaptation to the optimization of the location, and new ways of understanding the family, social relations, and the market.These theoreticians believe that the disappearance of the fordist models implies the proliferation of production spaces and decentralized work.The old factory, which required abundant labour, gives way to smaller companies with head offices distributed over different locations; part of the work is done, sent, or coordinated electronically; physical mobility and communication infrastructures make it possible to undo the residence and work in the same location binomial.A wide variety of theoretical traditions are included in these perspectives, such as the world systems theory (Timberlake), world city formation (Sassen), work in flexible production (Scott), and the information society (Castells).We could highlight the most explicit formulations on territorial dynamics in the work of Cheshire and Hay or Fielding.The most critical sociology is also based on these presumptions (Harvey), as are the broad approaches of rural and urban sociology (Wardwell, Hawley).
It is not surprising to find counterurban theoreticians incorporating their assumptions, for example, making low density places depend on preferences for life and work, located in turn in the structural changes that the economy was undergoing in general [36,40,43].Some clearly post-fordist authors, in turn, agreed with the counterurban assumptions, describing some cities as the remains of an industrial age when transport costs were too high, supply chains were local, and people lived close to their workplace; in the postindustrial world of low communication costs, people and companies prefer to be located in cheaper places, less congested and with greater environmental quality.A contrary point of view emerged later, identifying cities as centres of renewed economic dynamism and driving forces of national prosperity [2,11].
They all recognize the role of infrastructures, the reduction of transport time, and the importance of management and governance, together with structural changes in productive organization.They are separated by their consideration of emerging spaces.In general, however, they deny that the city or rural spaces have better or worse future options; on the contrary, the idea that potential is independent of dimension and location (urban, peripheral, or rural) is predominant, as opposed to another kind of variables related to insertion into networks (often extraterritorial and/or global), in which any territory has potential for local development.Only territories that are connected in these networks have the capacity for dynamism in this globalized space [44].In a way, postfordism is a temptation towards the empirical versatility to which any situation can adapt, whatever the balances and

"Many sparsely settled nonmetropolitan areas still depend on extractive industries continue to lose population, as they have for decades. Yet other thinly settled areas with similar histories of dependence on extractive industries are experiencing renewed growth as centers of recreation or retirement" [7, page 540].
The underlying logic that operates before this open spectrum of possibilities is urban management, or as it has been called more recently governance.In addition to the inertia inherent in productive restructuring, cities have been seen historically as immersed in deficient management, which explains both the turnaround and other processes of suburbanization, counterurbanization, and reurbanization."The two theories (native evolution of social systems and policies) have a number of interactions and interrelations, and consequently, it is difficult to distinguish between them empirically" [45, page 137].

Cyclical Interpretation.
Different schemes are included in the paradigm "urban cycles" that propose an arrangement of the process of urbanization by phases, and inside each phase they suggest predominant senses, magnitudes, and types of migration.They are often based on causalities included in theories of productive restructuring and at times in those of individual preferences.Nevertheless, such consideration can also be seen as an ad hoc cause to the extent that it understands that cities and other territories have their own structural dynamics, almost as if they were alive.The consideration of this interpretation as causal rests, therefore, on focusing on stages, the result of an inertia inherent in every habitat.Only secondly do external causes act as participating structures.
The first formulations are already to be found in studies by Patrick Geddes, Cities in evolution, in 1913 and by Lewis Mumford, The Culture of the Cities, in 1938.At the end of the 1940s the Theory of the Tidal Wave was a significant precedent, although somewhat focused on orthodox centripetal logic.In the 1960s references became more numerous, with studies by Birch, Borchet, Fooreston, and Wilson.Later; the formulation became more specific in the work of Hall, Klaassen, van den Berg, Champion, Cheshire, or Aydalot.More recently the model was discussed in Parr [12].We should also mention the applications and empirical contrasts of this interpretation in numerous other studies.
The standard model of urban cycles [25,46,47] implies an indefinite sequence of growth and decrease between the core and the ring.The basic, more general, and widely disseminated presentation is based on four phases or moments of growth: 1/urbanization, the process of population concentration in cities, also generating suburban rings to the detriment of rural territories; 2/deconcentration (exurbanization) when the urban ring grows at the expense of the urban core; 3/deurbanization or counterurbanization, when both the urban core and its ring lose population; 4/reurbanization, when the core changes again to a positive balance, while the ring loses population (Figure 2(a)).
The description of the process attempts to adapt to this model and the cause for this cyclical sequence happening is explained, very tacitly, by all interpretive possibilities (inefficiencies in governance, the cost of housing, rural residential preferences, and changes in productive structures) that is temporary and that leads to urban saturation (systemic maturation).The main problem or explanatory cause of the sequence is linked, nevertheless, to housing and in a triple sense: the size of dwelling, their price and the quality of building.A series of environmental problems have been added to this (noise, cleanliness, citizens' safety from crime, etc.) that make living in a given location less attractive (thus linking a causality of residential preferences).All this has led to postulating the idea that when a city or area grows to a certain level, at which it reaches a saturation point, processes of deconcentration and counterurbanization take place.Nevertheless, these "crises" end up by being favourably solved for cities, predicting, in general, processes of reurbanization.Alternatively, the cyclical model is complemented with that of the theoreticians of transition (Figure 2(b)), with the difference that the latter do not consider urban rebirth as hegemonic, but rather find balance among all habitats [35,40,48,49].
Champion [24] takes the hypothesis of urban cycles as a proven fact by observing the evolution of British urban areas in the second half of the twentieth century.Nevertheless, the shortcomings of the model are also evident.By way of example, the central idea that territories should follow a sequence of stages is questioned, at least partly, with examples that throughout the second half of the twentieth century have not shown signs of change (this is shown for Europe by Cheshire and Hay [46, page 145] and for Germany [26].On the other hand, the sequence followed is different in some cases.For example, for Holland, van der Berg, and van der Meer (1981, diag.10.1) point out that 25% of Dutch FURs skip one phase from 1950 to 1978.Hall and Hay [26,Table A.45] point out that 20% of German areas skip one and 30% skip more than one stage from 1950 to 1970.Reversions also take place, especially as reported by van den Berg et al. [50,Table 8.6] and are also noticeable in North American official statistics [51,Table 25]).Nyström concluded that while the first three phases are established processes, "it has only been possible to detect reurbanization in certain cities" [49, page 143].

Diagnosis
4.1.Empirical Processes.The evolution of population by habitat, both in America and in Europe, can be evaluated with greater precision over time.If at the beginning of the turnaround evidence was confused as it did not internally differentiate nonmetropolitan areas, the data prepared later enables solving some doubts.Suspicions were focused on the fact that these areas make reference, indistinctly, to urban and rural spaces and what is more significant to peripheral spaces in metropolitan areas (some of which would later become metropolitan).With Strauss [3] and Gordon [52], who analyze the counties of the nonmetropolitan peripheral areas separate from metropolitan areas, the trend of the dynamism of peripheries is visualized, concealed in a supposed counterurbanization.Long and Nucci [4] recalculate the growth and concentration indicators (particularly the Hoover index) for North America over the period 1960-1994, divided into three groups: counties that were metropolitan throughout the whole period, counties that were nonmetropolitan throughout, and counties that went from nonmetropolitan to metropolitan.Metropolitan counties do not evidence anything new with regard to what is already known; they grow in the 1960s, diminish their relative growth in the 1970s, grow again, although very slightly, in the 1980s and then reduce their growth, also very slightly, in the early 1990s.Nonmetropolitan counties show inverse trends for each period, but with one relevant exception: they did not grow in the 1990s.Finally, counties which were nonmetropolitan and became metropolitan always grow in each and every one of the decades under study.Their growth, furthermore, explains most of the entire growth in the whole second half of the century.Further studies refined the methodologies, internally differentiating nonmetropolitan areas [19] and, very specifically, identifying their "peripheral" nature [53].In the new millennium, a series of studies were published, demystifying counterurban theses even more clearly.The empirical analysis of interior migration data enabled the discovery of how the North American rebound in the early 1990s went backwards in the second half of the decade [5][6][7]54], measured in terms of net migration.
Wang's research [54] showed how, for the period 1975-1980, when nonmetropolitan areas gained over 2 million inhabitants in their interior migratory balance, 70.8% of them took place in counties "adjacent" to metropolitan areas.Furthermore, "residual" counties (those that do not form part of the adjacent ones, or those belonging to "functional urban regions") maintained a negative net population balance.The conclusions of this study indicate once again the same trend for the rural rebound period, calculating flows from 1995 to 2000.
The identification of new metropolitan counties-those that were initially nonmetropolitan, but due to the expansion of a metropolitan area are included therein-sums up the location of the growth in quite a significant way, Johnson and collaborators [7]. Figure 3 is sufficiently eloquent on the matter.In fact, it allows us to visualize the evolution of the (relative) population increase by type of habitat most clearly, from 1963 to 2003.Authors use the county as the unit of analysis.Counties are classified as metropolitan or nonmetropolitan using criteria developed by the U.S. Office of Management and Budget.If one of these is considered nonmetropolitan in 1963 and after becomes metropolitan, "new metropolitan will be designed." On the other hand, International Journal of Population Research those that are considered metropolitan permanently are subdivided into two groups: the "metropolitan core" counties were already densely settled by 1963 (they contain most of the nation's older and larger central cities as well as many older inner suburbs); "metropolitan noncore" counties that were already metropolitan by 1963, but were outside these metro cores.
The data presented by Johnson et al. [7] reflects that all North American habitats grew throughout the second half of the twentieth century.This growth, nevertheless, was unequal, depending on the urban nature thereof.Peripheries, in general, defined as "new metropolitan counties" and permanently metropolitan peripheral counties in the "core" (i.e., "metropolitan non-core") doubled their population over this period.On the contrary, the growth of traditional compact cities and rural spaces (metropolitan core and nonmetropolitan, resp.)slowed down or stagnated (cf. Figure 3).
According to this and other sources, deconcentration existed but was highly selective, at least in a significantly statistical way, centred only on peripheries.The latter, as distinctly indefinite and diffuse spaces, not only monopolized the majority of growth, but also expanded and extended territorially, greatly increasing the perimeter of the surface they include.
These processes of peripheral deconcentration follow processes of suburbanization sequentially, more characteristic of a previous phase.They differ, nevertheless, in that suburban growth is accompanied by growth in compact cities; on the contrary, deconcentration implies the existence of centrifugal migratory flows from the centre to the periphery, to the loss of the inner city.The coincidence of peripheral growths in both suburban and deconcentration process has induced confusions.This has led to similar considering both phenomena, generalizing it in space and time.However, their role is opposite: the suburbanization accompanies the growth of core city; the deconcentration cancels it.
More recently, inner cities have inverted the negative trends recorded during the deconcentration processes.This new phase has been called reurbanization.This has been shown empirically in numerous studies, such as Mercer [55], who calculates an increase in the number of county-cores with a positive balance of 55% in the 1990s, compared to 42% in the 1980s.The observation of symptoms of reurbanization, although not unanimous, has been repeatedly shown in numerous European studies.
However, at least from the North American perspective, the measurement of the balances for cities, as opposed to what is happening in the periphery, does not allow us to extract this conclusion so easily.In fact, the evolution of the population balance for core cities is still negative, although it showed a clearly positive trend, involuted with the crisis in 2007 and afterwards (cf. Figure 4).
In Europe trends are much more diverse than in the USA, as the European Environment Agency [56] has stated in accordance with previous balances [57].In an individualized analysis of the growth dynamics of European cities, Turok and Mykhnenko [11] synthesize the different patterns of behaviour for 310 cities in 36 different countries.Practically all of them were growing in 1960; the trend towards a reduced number of cities that were growing was constant but it was interrupted with the new millennium, thereby enabling support for the hypothesis of "urban revival" (see Figure 5).
Recent history has left sufficient proof of the existence of a spatially extended phenomenon of deconcentration, but also this has been selective and fundamentally limited to the emerging peripheral ring.This trend was disturbed by the incidence of local policies, such as the activation of tourism or the establishment of retirement areas [44] or other elements of economic activation or deactivation.The future is drawn with a series of more diverse prognosis, among which reurbanization and territorial transition seem to be located in the first positions.
In any case, the classical city still performs an active role and maintains the explanatory centre of dynamism in territorial structure.What has changed is the geographical scope to which the city refers, going from the traditional compact city to the urban area and exceeding the administrative divisions of the local scale.

Conclusions: Theoretical Balance. The emergence of inter-
pretations about what was happening with the territorial dynamics was initially very confusing, oscillating in opposite directions and sustained in analysis of habitat not enough relevant.The use of the categories "metropolitan areas" and "nonmetropolitan areas" in the US is a clear example of this.Recently, some studies have shown that the contiguity to the cities and to a lesser extent, urban areas, explained the observed demographic balances [7,19,54,55].In that sense, we can identify three significant spaces for understanding the dynamics of each region: cities, periphery, and rural.Growth, demographic stagnation, or regression can be interpreted in three different ways, including the causal interpretation and prognosis of future.These theories can be identified and named in the academic literature as counterurban, postfordist, and cyclical theories.The balance of these interpretative paradigms from the point of view of empirical evidence shows some shortcomings in each one but also offers certain potentialities.
The paradigm of counterurban deconcentration has come up against not only the prourban evidence seen in the 1980s and in the first decade of the twenty-first century (turnaround reversal and rebound reversal) but also its representative periods (the turnaround of the 1970s and the rural rebound of the 1990s) have been questioned and seen as "a statistical illusion" [27,58].In fact, rural areas were not growing, but only the peripheral spaces of traditional cities. Theoreticians of counterurbanization nevertheless maintain its validity, not so much to carry out a predominant prognosis on urban territories but rather to detect a permanent flow, although not hegemonic, towards rural areas [22,33,42,44].
The paradigm of post-fordist restructuring seems to be well moulded to empirically observed diversity, although two suspicions permanently gravitate thereon.On the one hand, because they seem to rest firstly on empirical evidence and later construct a theoretical model that adapts to it, a positive result is guaranteed with this sequence.On the other hand, any situation or dynamic can be considered successful under this theory, since it is always possible to find synergies, political action, or any other structural circumstance (belonging to the system) or determined product of social agents that could be taken as a causative argument for the effects found.Not only a certain methodological individualism but also a temptation to bet on the winning horse in advance and in all safety underlies this.Regardless of the ontological debate, the result, nevertheless, can be considered useful and acceptable.
Cyclical theories usually provide territories with life; they seem to enjoy both autonomy with regard to social agents (and their policies) and high doses of determinism that make their phases inevitable.This is their main debility.Cities, just like any other habitat, whether peripheral or rural, are not independent from the society they are inserted into and so the policies applied thereto (or omitted) lead to the results.Good governance influences the securing of better results both in population balances and in immigration flows.This does not prevent us from agreeing with this theoretical framework which states that the sequence that defines the basic scheme is broadly verifiable.
These three paradigms cannot explain all diversity that exists in urban systems and territorial dynamics such as polycentric urban regions or the urbanization of the suburbs, but it is possible to learn a lot from them.The balance of their contributions suggests that they may converge in a theoretical synthesis.Counterurban has shown the way towards the relevancy of certain decisional components and how the structural determines individual preferences.Postfordism has enabled us to understand that there is no such thing as condemned territories, but only territories with no strategic project.Cyclical theories admit the role of predominant structural inertia.Bringing together structural inertia, individual preferences and strategic planning do not seem to be an impossible exercise.The design of a theoretical framework that synthesizes existing contributions as an outstanding task is a necessary, feasible, and enriching task.

Figure 3 :
Figure 3: Population evolution in the USA by type of habitat (population in 1960 = 100).Source: prepared from information presented by Johnson et al. [7, page 533] who in turn used the populations estimated by the U.S. Bureau of Census.

Figure 4 :
Figure 4: Population balance in the USA.Source: own, based on the U.S. Census Bureau, current population survey.

Table 1 :
Schematic evolution of the dynamics of urban-rural growth.

Table 2 :
Summary of interpretations.

Table 3 :
Theoretical alternatives for the future evolution of territorial dynamism.